The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 18 Jul 2012 15:08:40 +0300

List of important changes from nfse-20120310 to nfse-20120718.

Command line options -2, -n and -r now are not supported by nfse.

Added new command line option -m to nfse, that specifies whether nfse
should service MOUNT protocol requests and what it should support from
this protocol.

Added new option -resvport to nfs.exports, that specifies whether NFS and
NLM RPC requests should be received from sender's reserved IP port numbers
(default, always and weakauth modes).

Added new option -mnt_mount to nfs.exports, that specifies settings for
MOUNT protocol's procedures MNT, UMNT and UMNTALL (default, regfile and
resvport arguments).

Tried to make the nfse compatible mode completely compatible with
exports(5) configurations: http://nfse.sourceforge.net/COMPATIBILITY
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 18 Jul 2012 15:08:40 +0300

List of important changes from nfse-20120310 to nfse-20120718.

Command line options -2, -n and -r now are not supported by nfse.

Added new command line option -m to nfse, that specifies whether nfse
should service MOUNT protocol requests and what it should support from
this protocol.

Added new option -resvport to nfs.exports, that specifies whether NFS and
NLM RPC requests should be received from sender's reserved IP port numbers
(default, always and weakauth modes).

Added new option -mnt_mount to nfs.exports, that specifies settings for
MOUNT protocol's procedures MNT, UMNT and UMNTALL (default, regfile and
resvport arguments).

Tried to make the nfse compatible mode completely compatible with
exports(5) configurations: http://nfse.sourceforge.net/COMPATIBILITY
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sat, 20 Oct 2012 11:43:17 +0200

Andrey,

I'd really like to use this. However, I need to use it with FreeBSD 7.4,
8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
changes, but this turned out to be too difficult for me.

I believe your work could be more easily adopted (even into the core
FreeBSD sources) provided that
- patches for all supported branches of FreeBSD were available and
- there existed a simple knob in rc.conf where one switches between
the old mountd and the new nfse

I guess that for the latter you'd also need to introduce some
compatibility shims into your kernel changes, such that a single kernel
could support both methods.

What is your opinion?

Regards,

Martin

p.s. I have seen that just a few days a workaround was committed to
suspend the nfs daemons while mountd reloads the exports list, but this
is just a stopgap (as mentioned in the commit message). I'd much more
prefer nfse.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 18 Jul 2012 15:08:40 +0300

List of important changes from nfse-20120310 to nfse-20120718.

Command line options -2, -n and -r now are not supported by nfse.

Added new command line option -m to nfse, that specifies whether nfse
should service MOUNT protocol requests and what it should support from
this protocol.

Added new option -resvport to nfs.exports, that specifies whether NFS and
NLM RPC requests should be received from sender's reserved IP port numbers
(default, always and weakauth modes).

Added new option -mnt_mount to nfs.exports, that specifies settings for
MOUNT protocol's procedures MNT, UMNT and UMNTALL (default, regfile and
resvport arguments).

Tried to make the nfse compatible mode completely compatible with
exports(5) configurations: http://nfse.sourceforge.net/COMPATIBILITY
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sat, 20 Oct 2012 11:43:17 +0200

Andrey,

I'd really like to use this. However, I need to use it with FreeBSD 7.4,
8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
changes, but this turned out to be too difficult for me.

I believe your work could be more easily adopted (even into the core
FreeBSD sources) provided that
- patches for all supported branches of FreeBSD were available and
- there existed a simple knob in rc.conf where one switches between
the old mountd and the new nfse

I guess that for the latter you'd also need to introduce some
compatibility shims into your kernel changes, such that a single kernel
could support both methods.

What is your opinion?

Regards,

Martin

p.s. I have seen that just a few days a workaround was committed to
suspend the nfs daemons while mountd reloads the exports list, but this
is just a stopgap (as mentioned in the commit message). I'd much more
prefer nfse.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 22 Oct 2012 17:35:52 +0300

On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
> Andrey,
>
> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
> changes, but this turned out to be too difficult for me.

The nfse utility could be built on 8.x, 9.x and 10.x. To build nfse on 7.x
it is necessary to copy with LIST_SWAP() macro definition.

The semantics of the getgroupslist(3) was changed in 8.x and 9.x, so it is
required to use -mapall and -maproot with list of groups (-mapall=user:group)
on 7.x, otherwise nfse will report about duplicated groups for "-mapall=user".

The src/sys.diff changes are more complex. First of all NFSv4 support
and sys/fs/nfs* were added in 8.0 and some parts in sys/nfs* were changed.
Having checked difference between 7.x and 8.x-10.x in NFS related parts,
I think it will require to rewrite all changes from sys.diff to support 7.x.

Right now it is possible to apply sys.diff to RELENG_9, only one simple
change will be rejected. I applied sys.diff to 8.2 and there are several
rejected files, I made sys.diff for 8.2 and sent it to you in a private
message (it can be built, but I did not run it). I did not make changes
for cddl/ from 8.2, this is not necessary. Modification of etc/ is not
complex.

>
> I believe your work could be more easily adopted (even into the core
> FreeBSD sources) provided that
> - patches for all supported branches of FreeBSD were available and

I used to think that major changes should go to the CURRENT and only
then be MFCed to other releases if necessary.

> - there existed a simple knob in rc.conf where one switches between
> the old mountd and the new nfse

It works like this. Set mountd_enable="NO" and set nfse_enable="YES",
if compatibility mode with exports(5) configuration is required then
set nfse_exports_compat="YES".

>
> I guess that for the latter you'd also need to introduce some
> compatibility shims into your kernel changes, such that a single kernel
> could support both methods.

Does "Both methods" mean "mountd and nfse"? If so, then it works like
this right now. By default VFS_CHECKEXP() is used (mountd mode) and if
nfse is called at least one time, then NFSE code in the NFS server is used.
A user can continue to use mountd on modified system.

>
> What is your opinion?
>

There is opinion that nfse in its compatibility mode is not compatible
with mountd and its configuration. This opinion exist, because first
versions of NFSE changes were incompatible with existent exports(5)
configurations and I explained to several commiters why exports(5) format
is wrong and why nfs.exports(5) format was created. Then the compatibility
mode was implemented and then it was improved several times. I wrote
comparison of the nfse compatibility mode with mountd and exports(5):

http://nfse.sourceforge.net/COMPATIBILITY

According to my understanding of exports(5) format I can say that "nfse -C"
is completely compatible with exports(5) rules and mountd is incompatible
with exports(5) rules. Please note, that I do not compare nfs.exports(5)
with exports(5), because they are not compatible. The compatibility mode
was implemented because I was told that supporting of existent exports(5)
configuration is required.

Also I made several changes to mountd to make it more compatible with
exports(5) (available in PRs), or at least restore its compatibility
with exports(5) that existed several years ago. I asked questions whether
these changes are POLA violations here:

http://lists.freebsd.org/pipermail/freebsd-fs/2012-September/015175.html

I have not received any replies on my requests to check nfse compatibility
mode with existent configurations, so I think there is no opinion in the
community about nfse compatibility mode with exports(5) configuration.
Also, I have not received any replies about nfs.exports(5) configuration
format and any technical opinions/review about NFSE changes in the kernel.

I will repeat it again one more time. It is not necessary to modify
the kernel, patch or install something to check nfse compatibility mode
and/or to check nfs.exports(5) format, just compile the nfse utility and
use it to verify existent exports(5) configurations.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 18 Jul 2012 15:08:40 +0300

List of important changes from nfse-20120310 to nfse-20120718.

Command line options -2, -n and -r now are not supported by nfse.

Added new command line option -m to nfse, that specifies whether nfse
should service MOUNT protocol requests and what it should support from
this protocol.

Added new option -resvport to nfs.exports, that specifies whether NFS and
NLM RPC requests should be received from sender's reserved IP port numbers
(default, always and weakauth modes).

Added new option -mnt_mount to nfs.exports, that specifies settings for
MOUNT protocol's procedures MNT, UMNT and UMNTALL (default, regfile and
resvport arguments).

Tried to make the nfse compatible mode completely compatible with
exports(5) configurations: http://nfse.sourceforge.net/COMPATIBILITY
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sat, 20 Oct 2012 11:43:17 +0200

Andrey,

I'd really like to use this. However, I need to use it with FreeBSD 7.4,
8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
changes, but this turned out to be too difficult for me.

I believe your work could be more easily adopted (even into the core
FreeBSD sources) provided that
- patches for all supported branches of FreeBSD were available and
- there existed a simple knob in rc.conf where one switches between
the old mountd and the new nfse

I guess that for the latter you'd also need to introduce some
compatibility shims into your kernel changes, such that a single kernel
could support both methods.

What is your opinion?

Regards,

Martin

p.s. I have seen that just a few days a workaround was committed to
suspend the nfs daemons while mountd reloads the exports list, but this
is just a stopgap (as mentioned in the commit message). I'd much more
prefer nfse.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 22 Oct 2012 17:35:52 +0300

On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
> Andrey,
>
> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
> changes, but this turned out to be too difficult for me.

The nfse utility could be built on 8.x, 9.x and 10.x. To build nfse on 7.x
it is necessary to copy with LIST_SWAP() macro definition.

The semantics of the getgroupslist(3) was changed in 8.x and 9.x, so it is
required to use -mapall and -maproot with list of groups (-mapall=user:group)
on 7.x, otherwise nfse will report about duplicated groups for "-mapall=user".

The src/sys.diff changes are more complex. First of all NFSv4 support
and sys/fs/nfs* were added in 8.0 and some parts in sys/nfs* were changed.
Having checked difference between 7.x and 8.x-10.x in NFS related parts,
I think it will require to rewrite all changes from sys.diff to support 7.x.

Right now it is possible to apply sys.diff to RELENG_9, only one simple
change will be rejected. I applied sys.diff to 8.2 and there are several
rejected files, I made sys.diff for 8.2 and sent it to you in a private
message (it can be built, but I did not run it). I did not make changes
for cddl/ from 8.2, this is not necessary. Modification of etc/ is not
complex.

>
> I believe your work could be more easily adopted (even into the core
> FreeBSD sources) provided that
> - patches for all supported branches of FreeBSD were available and

I used to think that major changes should go to the CURRENT and only
then be MFCed to other releases if necessary.

> - there existed a simple knob in rc.conf where one switches between
> the old mountd and the new nfse

It works like this. Set mountd_enable="NO" and set nfse_enable="YES",
if compatibility mode with exports(5) configuration is required then
set nfse_exports_compat="YES".

>
> I guess that for the latter you'd also need to introduce some
> compatibility shims into your kernel changes, such that a single kernel
> could support both methods.

Does "Both methods" mean "mountd and nfse"? If so, then it works like
this right now. By default VFS_CHECKEXP() is used (mountd mode) and if
nfse is called at least one time, then NFSE code in the NFS server is used.
A user can continue to use mountd on modified system.

>
> What is your opinion?
>

There is opinion that nfse in its compatibility mode is not compatible
with mountd and its configuration. This opinion exist, because first
versions of NFSE changes were incompatible with existent exports(5)
configurations and I explained to several commiters why exports(5) format
is wrong and why nfs.exports(5) format was created. Then the compatibility
mode was implemented and then it was improved several times. I wrote
comparison of the nfse compatibility mode with mountd and exports(5):

http://nfse.sourceforge.net/COMPATIBILITY

According to my understanding of exports(5) format I can say that "nfse -C"
is completely compatible with exports(5) rules and mountd is incompatible
with exports(5) rules. Please note, that I do not compare nfs.exports(5)
with exports(5), because they are not compatible. The compatibility mode
was implemented because I was told that supporting of existent exports(5)
configuration is required.

Also I made several changes to mountd to make it more compatible with
exports(5) (available in PRs), or at least restore its compatibility
with exports(5) that existed several years ago. I asked questions whether
these changes are POLA violations here:

http://lists.freebsd.org/pipermail/freebsd-fs/2012-September/015175.html

I have not received any replies on my requests to check nfse compatibility
mode with existent configurations, so I think there is no opinion in the
community about nfse compatibility mode with exports(5) configuration.
Also, I have not received any replies about nfs.exports(5) configuration
format and any technical opinions/review about NFSE changes in the kernel.

I will repeat it again one more time. It is not necessary to modify
the kernel, patch or install something to check nfse compatibility mode
and/or to check nfs.exports(5) format, just compile the nfse utility and
use it to verify existent exports(5) configurations.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: Andrey Simonenko <.ua>, bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Fri, 26 Oct 2012 15:15:56 +0200

Hi Andrey,

Today I started applying your changes and did the following:

1. downloaded nfse-20121025.tar.bz2 from sourceforge
2. read INSTALL-all
3. checked out release/8.2.0 from FreeBSD SVN
4. applied src/cddl.diff
==> this failed
5. checked out head from FreeBSD SVN
6. applied src/cddl diff
==> this failed as well

I have imported all nfse patch files from sourceforge in a local
mercurial repo to be able to easier follow what is changing. There I see
that cddl.diff was updated for the last time on May 17.

Could you help me with the following questions:
- Is INSTALL-all still relevant, and if yes, for which cases?
- What for is cddl.diff?
- I am heavily using zfs. Which patches from your patchset do I need to
get nfse to fully support zfs?

Lastly, I believe it might be more helpful to combine INSTALL-all and
INSTALL-kern into a single file INSTALL and in that file clearly point
out the differences between the two methods (what does one method give
you, what the other, what do I need to do for the first method, what for
the other).

Regards,

Martin

On 10/22/12 16:37, Andrey Simonenko wrote:
> On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
>> Andrey,
>>
>> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
>> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
>> changes, but this turned out to be too difficult for me.
>>
> sys.diff for 8.2 in the attachment.

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 18 Jul 2012 15:08:40 +0300

List of important changes from nfse-20120310 to nfse-20120718.

Command line options -2, -n and -r now are not supported by nfse.

Added new command line option -m to nfse, that specifies whether nfse
should service MOUNT protocol requests and what it should support from
this protocol.

Added new option -resvport to nfs.exports, that specifies whether NFS and
NLM RPC requests should be received from sender's reserved IP port numbers
(default, always and weakauth modes).

Added new option -mnt_mount to nfs.exports, that specifies settings for
MOUNT protocol's procedures MNT, UMNT and UMNTALL (default, regfile and
resvport arguments).

Tried to make the nfse compatible mode completely compatible with
exports(5) configurations: http://nfse.sourceforge.net/COMPATIBILITY
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sat, 20 Oct 2012 11:43:17 +0200

Andrey,

I'd really like to use this. However, I need to use it with FreeBSD 7.4,
8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
changes, but this turned out to be too difficult for me.

I believe your work could be more easily adopted (even into the core
FreeBSD sources) provided that
- patches for all supported branches of FreeBSD were available and
- there existed a simple knob in rc.conf where one switches between
the old mountd and the new nfse

I guess that for the latter you'd also need to introduce some
compatibility shims into your kernel changes, such that a single kernel
could support both methods.

What is your opinion?

Regards,

Martin

p.s. I have seen that just a few days a workaround was committed to
suspend the nfs daemons while mountd reloads the exports list, but this
is just a stopgap (as mentioned in the commit message). I'd much more
prefer nfse.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 22 Oct 2012 17:35:52 +0300

On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
> Andrey,
>
> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
> changes, but this turned out to be too difficult for me.

The nfse utility could be built on 8.x, 9.x and 10.x. To build nfse on 7.x
it is necessary to copy with LIST_SWAP() macro definition.

The semantics of the getgroupslist(3) was changed in 8.x and 9.x, so it is
required to use -mapall and -maproot with list of groups (-mapall=user:group)
on 7.x, otherwise nfse will report about duplicated groups for "-mapall=user".

The src/sys.diff changes are more complex. First of all NFSv4 support
and sys/fs/nfs* were added in 8.0 and some parts in sys/nfs* were changed.
Having checked difference between 7.x and 8.x-10.x in NFS related parts,
I think it will require to rewrite all changes from sys.diff to support 7.x.

Right now it is possible to apply sys.diff to RELENG_9, only one simple
change will be rejected. I applied sys.diff to 8.2 and there are several
rejected files, I made sys.diff for 8.2 and sent it to you in a private
message (it can be built, but I did not run it). I did not make changes
for cddl/ from 8.2, this is not necessary. Modification of etc/ is not
complex.

>
> I believe your work could be more easily adopted (even into the core
> FreeBSD sources) provided that
> - patches for all supported branches of FreeBSD were available and

I used to think that major changes should go to the CURRENT and only
then be MFCed to other releases if necessary.

> - there existed a simple knob in rc.conf where one switches between
> the old mountd and the new nfse

It works like this. Set mountd_enable="NO" and set nfse_enable="YES",
if compatibility mode with exports(5) configuration is required then
set nfse_exports_compat="YES".

>
> I guess that for the latter you'd also need to introduce some
> compatibility shims into your kernel changes, such that a single kernel
> could support both methods.

Does "Both methods" mean "mountd and nfse"? If so, then it works like
this right now. By default VFS_CHECKEXP() is used (mountd mode) and if
nfse is called at least one time, then NFSE code in the NFS server is used.
A user can continue to use mountd on modified system.

>
> What is your opinion?
>

There is opinion that nfse in its compatibility mode is not compatible
with mountd and its configuration. This opinion exist, because first
versions of NFSE changes were incompatible with existent exports(5)
configurations and I explained to several commiters why exports(5) format
is wrong and why nfs.exports(5) format was created. Then the compatibility
mode was implemented and then it was improved several times. I wrote
comparison of the nfse compatibility mode with mountd and exports(5):

http://nfse.sourceforge.net/COMPATIBILITY

According to my understanding of exports(5) format I can say that "nfse -C"
is completely compatible with exports(5) rules and mountd is incompatible
with exports(5) rules. Please note, that I do not compare nfs.exports(5)
with exports(5), because they are not compatible. The compatibility mode
was implemented because I was told that supporting of existent exports(5)
configuration is required.

Also I made several changes to mountd to make it more compatible with
exports(5) (available in PRs), or at least restore its compatibility
with exports(5) that existed several years ago. I asked questions whether
these changes are POLA violations here:

http://lists.freebsd.org/pipermail/freebsd-fs/2012-September/015175.html

I have not received any replies on my requests to check nfse compatibility
mode with existent configurations, so I think there is no opinion in the
community about nfse compatibility mode with exports(5) configuration.
Also, I have not received any replies about nfs.exports(5) configuration
format and any technical opinions/review about NFSE changes in the kernel.

I will repeat it again one more time. It is not necessary to modify
the kernel, patch or install something to check nfse compatibility mode
and/or to check nfs.exports(5) format, just compile the nfse utility and
use it to verify existent exports(5) configurations.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: Andrey Simonenko <.ua>, bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Fri, 26 Oct 2012 15:15:56 +0200

Hi Andrey,

Today I started applying your changes and did the following:

1. downloaded nfse-20121025.tar.bz2 from sourceforge
2. read INSTALL-all
3. checked out release/8.2.0 from FreeBSD SVN
4. applied src/cddl.diff
==> this failed
5. checked out head from FreeBSD SVN
6. applied src/cddl diff
==> this failed as well

I have imported all nfse patch files from sourceforge in a local
mercurial repo to be able to easier follow what is changing. There I see
that cddl.diff was updated for the last time on May 17.

Could you help me with the following questions:
- Is INSTALL-all still relevant, and if yes, for which cases?
- What for is cddl.diff?
- I am heavily using zfs. Which patches from your patchset do I need to
get nfse to fully support zfs?

Lastly, I believe it might be more helpful to combine INSTALL-all and
INSTALL-kern into a single file INSTALL and in that file clearly point
out the differences between the two methods (what does one method give
you, what the other, what do I need to do for the first method, what for
the other).

Regards,

Martin

On 10/22/12 16:37, Andrey Simonenko wrote:
> On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
>> Andrey,
>>
>> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
>> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
>> changes, but this turned out to be too difficult for me.
>>
> sys.diff for 8.2 in the attachment.

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 29 Oct 2012 13:55:42 +0200

Hello,

On Fri, Oct 26, 2012 at 03:15:56PM +0200, Martin Birgmeier wrote:
> Hi Andrey,
>
> Today I started applying your changes and did the following:
>
> 1. downloaded nfse-20121025.tar.bz2 from sourceforge
> 2. read INSTALL-all
> 3. checked out release/8.2.0 from FreeBSD SVN
> 4. applied src/cddl.diff
> ==> this failed

The cddl.diff file cannot be applied to cddl/ from 8.2, since cddl/ source
was updated several times in next FreeBSD versions. These changes are not
strictly necessary (see below).

> 5. checked out head from FreeBSD SVN
> 6. applied src/cddl diff
> ==> this failed as well

Check again, I can apply cddl.diff to just csup'ed 10-CURRENT.

>
> I have imported all nfse patch files from sourceforge in a local
> mercurial repo to be able to easier follow what is changing. There I see
> that cddl.diff was updated for the last time on May 17.
>
> Could you help me with the following questions:
> - Is INSTALL-all still relevant, and if yes, for which cases?

This file describes how to apply all NFSE changes to the FreeBSD source
code to make complete integration.

> - What for is cddl.diff?
> - I am heavily using zfs. Which patches from your patchset do I need to
> get nfse to fully support zfs?

This file contains integration of NFSE with the zfs program. When one
calls 'zfs sharenfs/unshare ...', then NFS exports settings are updated
using SIGHUP or dynamic NFSE commands (depends on presence of the
/etc/nfs.exports file). If you do not use NFSE dynamic commands
(eg. "nfse -c 'flush/clear/add/update/delete/set/unset ...'"), then
cddl.diff is not needed, just create symlink to mountd.pid by setting
the nfse_mountd_pid rc variable to "YES" and 'zfs share/unshare ...'
will send SIGHUP to nfse.

There is one question about NFSE and ZFS, this is support of ZFS snapshots.
NFSE was implemented as part of NFS server, not the part of VFS framework.
As a result right now it is impossible to automatically (unconditionally,
as it is done in all FreeBSD versions) export ZFS snapshots by NFSE.

>
> Lastly, I believe it might be more helpful to combine INSTALL-all and
> INSTALL-kern into a single file INSTALL and in that file clearly point
> out the differences between the two methods (what does one method give
> you, what the other, what do I need to do for the first method, what for
> the other).

You are right, I've just updated these two files with descriptions.

I suggest to apply changes to sys/ (I sent to you before) and etc/
(necessary to correct two rejected updates). Then build and install
the kernel and try to run nfse with simple configuration (/etc/exports
or /etc/nfs.exports).
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)

The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 14 Mar 2012 14:23:56 +0200

List of important changes from nfse-20100507 to nfse-20120310.

Improved integration with /etc/rc.d, improved compatibility mode
with mountd(8) ("nfse -C ..."), changes for the 10-CURRENT kernel.

Now specifications (credentials and security flavors) for export entities
(host or network) are given by array of user specified size. This change
to NFSE API allows to change number and types of specifications without
changing kernel ABI.

Added new NFSE_CMD_SPECS command to NFSE API that allows to load
specifications (credentials and security flavors) separately from export
settings.

Several options were converted to boolean datatype and now can be set
independently for each address specification, they are -nfsv2, -nfsv3,
-nfsv4, -mnt_dump and -mnt_export (boolean values "yes" and "no").

Added two new commands "set" and "unset" to nfse(8), these commands
are relaxed versions of "add/update" and "delete" commands.

Added new option -mnt_export_brief to nfs.exports(5).

Added nfse(8) support to zfs(8), if /etc/nfs.exports exists, then
"zfs sharenfs ..." and "zfs unshare ..." use nfse(8) commands.

URL: http://nfse.sourceforge.net/
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sun, 20 May 2012 10:04:01 +0200

Dear Andrey,

It seems that you have done some great work here, and I would really
like to see this integrated into the core FreeBSD distribution (I was
the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).

I would like to try out your patches and have two questions:

- Do your patches support multiple zfs sharenfs specifications as
proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
using this)?

- Could you give a concise list of incompatibilities (and even
regressions if they should exist at all) of your solution compared to
the standard one? - As to the advantages, I am already convinced. :-)

Thank you & regards,

Martin

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hello,

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
> The following reply was made to PR kern/136865; it has been noted by GNATS.
>
> From: Martin Birgmeier <>
> To: bug-, .ua
> Cc:
> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
> atomic updates
> Date: Sun, 20 May 2012 10:04:01 +0200
>
> Dear Andrey,
>
> It seems that you have done some great work here, and I would really
> like to see this integrated into the core FreeBSD distribution (I was
> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>
> I would like to try out your patches and have two questions:
>
> - Do your patches support multiple zfs sharenfs specifications as
> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
> using this)?

The exports(5) manual page says that address specifications must be specified
after options. The nfs.exports(5) file format allows to use options after
address specifications, so they can overwrite previously specified options.

It is possible to specify all settings for one file system in one line,
no ';' like separators are required.

For example line:

/fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3

will be translated to ("nfse -t ..." output):

Pathname /fs
Export specifications:
-rw -sec sys -maproot=-2:-2 -host 3.3.3.3
-rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
-ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1

>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)

In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
then it is more compatible with exports(5) than mountd is. If one did
not follow rules of exports(5), then "nfse -C ..." can be incompatible
with mountd.

If nfse is run in native nfs.export(5) configuration file format mode,
then logic of configuration looks like exports(5), but differs in some
places.

So, when we speak about "incompatibilities" then it is necessary to
distinguish incompatibilities of "nfse native mode" vs mountd and
incompatibilities of "nfse compatible mode" vs mountd.

I suggest to check whether "nfse -C ..." is compatible with mountd
using instructions described here:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html

You do not need to install anything or modify existent system for
testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
is compatible enough with mountd (try correct configurations and
configurations with mistakes).

I have list of difference somewhere, I'll try to find it.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
Hi Andrey,

One more question: I am running 8.3, 9.0, and 7.4 on various machines.
Do you have patches for these versions, too?

Regards,

Martin

On 05/22/12 10:04, Andrey Simonenko wrote:
> Hello,
>
> On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>> The following reply was made to PR kern/136865; it has been noted by GNATS.
>>
>> From: Martin Birgmeier<>
>> To: bug-, .ua
>> Cc:
>> Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
>> atomic updates
>> Date: Sun, 20 May 2012 10:04:01 +0200
>>
>> Dear Andrey,
>>
>> It seems that you have done some great work here, and I would really
>> like to see this integrated into the core FreeBSD distribution (I was
>> the submitter of http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/131342).
>>
>> I would like to try out your patches and have two questions:
>>
>> - Do your patches support multiple zfs sharenfs specifications as
>> proposed in http://www.freebsd.org/cgi/query-pr.cgi?pr=147881 (I am
>> using this)?
> The exports(5) manual page says that address specifications must be specified
> after options. The nfs.exports(5) file format allows to use options after
> address specifications, so they can overwrite previously specified options.
>
> It is possible to specify all settings for one file system in one line,
> no ';' like separators are required.
>
> For example line:
>
> /fs -ro -sec krb5 1.1.1.1 -nfsv4 no -rw 2.2.2.2 -sec sys -nfsv4 yes 3.3.3.3
>
> will be translated to ("nfse -t ..." output):
>
> Pathname /fs
> Export specifications:
> -rw -sec sys -maproot=-2:-2 -host 3.3.3.3
> -rw -sec krb5 -maproot=-2:-2 -nfsv4 no -host 2.2.2.2
> -ro -sec krb5 -maproot=-2:-2 -host 1.1.1.1
>
>>
>> - Could you give a concise list of incompatibilities (and even
>> regressions if they should exist at all) of your solution compared to
>> the standard one? - As to the advantages, I am already convinced. :-)
> In short: if nfse is run in compatible mode with mountd ("nfse -C ..."),
> then it is more compatible with exports(5) than mountd is. If one did
> not follow rules of exports(5), then "nfse -C ..." can be incompatible
> with mountd.
>
> If nfse is run in native nfs.export(5) configuration file format mode,
> then logic of configuration looks like exports(5), but differs in some
> places.
>
> So, when we speak about "incompatibilities" then it is necessary to
> distinguish incompatibilities of "nfse native mode" vs mountd and
> incompatibilities of "nfse compatible mode" vs mountd.
>
> I suggest to check whether "nfse -C ..." is compatible with mountd
> using instructions described here:
>
> http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
>
> You do not need to install anything or modify existent system for
> testing. Can you try "nfse -Ct ..." and tell me whether "nfse -C ..."
> is compatible enough with mountd (try correct configurations and
> configurations with mistakes).
>
> I have list of difference somewhere, I'll try to find it.
>
>
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 27, 2012 at 01:58:58PM +0200, Martin Birgmeier wrote:
>
> One more question: I am running 8.3, 9.0, and 7.4 on various machines.
> Do you have patches for these versions, too?

I just try to keep my changes for sys/ in sync with the CURRENT version only.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: freebsd-, bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 11 Jun 2012 13:16:41 +0300

On Sun, May 20, 2012 at 08:10:04AM +0000, Martin Birgmeier wrote:
>
> - Could you give a concise list of incompatibilities (and even
> regressions if they should exist at all) of your solution compared to
> the standard one? - As to the advantages, I am already convinced. :-)
>

This is the list of difference between "nfse -C ..." (compatible mode
with mountd) and mountd:

1. All options from exports(5) are supported, except obsolete options
-o, -n and -m (exports(5) mentions only -o for r/o access) and
the -webnfs option.

2. The -alldirs option works as described in exports(5). Right now
mountd ignores this option (it was broken almost ~5 years ago).

3. The -sec option works correctly (mountd incorrectly gives a remote
system information about security flavors).

4. If some option overwrites previous option, then new option will be used
Eg. "/fs host1 -ro host2", host1 will have r/w and host2 will have r/o
access mode.

5. If some pathname has wrong option, duplicated address specification,
then nfse will ignore all settings for this pathname.

6. If a line is not started with pathname, then nfse will ignore entire
configuration.

7. The -index option for the new NFS server will be ignored, since its
implementation in new and old NFS servers are different.

8. If several pathnames are given in one line, then they are not required
to belong to the same file system.

9. If the -h command line option is used, then nfse will require the -p
command line option as well.

10. The -2, -e, -n, -o, -r command line options are not supported, there
are better solutions as command line options or configuration options.

The list of differences between nfs.exports(5) and exports(5) is not
given, since it will require to copy here entire documentation.

Anyone can test whether nfse understands existent configurations without
installing nfse and recompiling the kernel using instructions from this
message:

http://lists.freebsd.org/pipermail/freebsd-fs/2010-May/008421.html
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Wed, 18 Jul 2012 15:08:40 +0300

List of important changes from nfse-20120310 to nfse-20120718.

Command line options -2, -n and -r now are not supported by nfse.

Added new command line option -m to nfse, that specifies whether nfse
should service MOUNT protocol requests and what it should support from
this protocol.

Added new option -resvport to nfs.exports, that specifies whether NFS and
NLM RPC requests should be received from sender's reserved IP port numbers
(default, always and weakauth modes).

Added new option -mnt_mount to nfs.exports, that specifies settings for
MOUNT protocol's procedures MNT, UMNT and UMNTALL (default, regfile and
resvport arguments).

Tried to make the nfse compatible mode completely compatible with
exports(5) configurations: http://nfse.sourceforge.net/COMPATIBILITY
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: bug-, .ua
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Sat, 20 Oct 2012 11:43:17 +0200

Andrey,

I'd really like to use this. However, I need to use it with FreeBSD 7.4,
8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
changes, but this turned out to be too difficult for me.

I believe your work could be more easily adopted (even into the core
FreeBSD sources) provided that
- patches for all supported branches of FreeBSD were available and
- there existed a simple knob in rc.conf where one switches between
the old mountd and the new nfse

I guess that for the latter you'd also need to introduce some
compatibility shims into your kernel changes, such that a single kernel
could support both methods.

What is your opinion?

Regards,

Martin

p.s. I have seen that just a few days a workaround was committed to
suspend the nfs daemons while mountd reloads the exports list, but this
is just a stopgap (as mentioned in the commit message). I'd much more
prefer nfse.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 22 Oct 2012 17:35:52 +0300

On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
> Andrey,
>
> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
> changes, but this turned out to be too difficult for me.

The nfse utility could be built on 8.x, 9.x and 10.x. To build nfse on 7.x
it is necessary to copy with LIST_SWAP() macro definition.

The semantics of the getgroupslist(3) was changed in 8.x and 9.x, so it is
required to use -mapall and -maproot with list of groups (-mapall=user:group)
on 7.x, otherwise nfse will report about duplicated groups for "-mapall=user".

The src/sys.diff changes are more complex. First of all NFSv4 support
and sys/fs/nfs* were added in 8.0 and some parts in sys/nfs* were changed.
Having checked difference between 7.x and 8.x-10.x in NFS related parts,
I think it will require to rewrite all changes from sys.diff to support 7.x.

Right now it is possible to apply sys.diff to RELENG_9, only one simple
change will be rejected. I applied sys.diff to 8.2 and there are several
rejected files, I made sys.diff for 8.2 and sent it to you in a private
message (it can be built, but I did not run it). I did not make changes
for cddl/ from 8.2, this is not necessary. Modification of etc/ is not
complex.

>
> I believe your work could be more easily adopted (even into the core
> FreeBSD sources) provided that
> - patches for all supported branches of FreeBSD were available and

I used to think that major changes should go to the CURRENT and only
then be MFCed to other releases if necessary.

> - there existed a simple knob in rc.conf where one switches between
> the old mountd and the new nfse

It works like this. Set mountd_enable="NO" and set nfse_enable="YES",
if compatibility mode with exports(5) configuration is required then
set nfse_exports_compat="YES".

>
> I guess that for the latter you'd also need to introduce some
> compatibility shims into your kernel changes, such that a single kernel
> could support both methods.

Does "Both methods" mean "mountd and nfse"? If so, then it works like
this right now. By default VFS_CHECKEXP() is used (mountd mode) and if
nfse is called at least one time, then NFSE code in the NFS server is used.
A user can continue to use mountd on modified system.

>
> What is your opinion?
>

There is opinion that nfse in its compatibility mode is not compatible
with mountd and its configuration. This opinion exist, because first
versions of NFSE changes were incompatible with existent exports(5)
configurations and I explained to several commiters why exports(5) format
is wrong and why nfs.exports(5) format was created. Then the compatibility
mode was implemented and then it was improved several times. I wrote
comparison of the nfse compatibility mode with mountd and exports(5):

http://nfse.sourceforge.net/COMPATIBILITY

According to my understanding of exports(5) format I can say that "nfse -C"
is completely compatible with exports(5) rules and mountd is incompatible
with exports(5) rules. Please note, that I do not compare nfs.exports(5)
with exports(5), because they are not compatible. The compatibility mode
was implemented because I was told that supporting of existent exports(5)
configuration is required.

Also I made several changes to mountd to make it more compatible with
exports(5) (available in PRs), or at least restore its compatibility
with exports(5) that existed several years ago. I asked questions whether
these changes are POLA violations here:

http://lists.freebsd.org/pipermail/freebsd-fs/2012-September/015175.html

I have not received any replies on my requests to check nfse compatibility
mode with existent configurations, so I think there is no opinion in the
community about nfse compatibility mode with exports(5) configuration.
Also, I have not received any replies about nfs.exports(5) configuration
format and any technical opinions/review about NFSE changes in the kernel.

I will repeat it again one more time. It is not necessary to modify
the kernel, patch or install something to check nfse compatibility mode
and/or to check nfs.exports(5) format, just compile the nfse utility and
use it to verify existent exports(5) configurations.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Martin Birgmeier <>
To: Andrey Simonenko <.ua>, bug-
Cc:
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Fri, 26 Oct 2012 15:15:56 +0200

Hi Andrey,

Today I started applying your changes and did the following:

1. downloaded nfse-20121025.tar.bz2 from sourceforge
2. read INSTALL-all
3. checked out release/8.2.0 from FreeBSD SVN
4. applied src/cddl.diff
==> this failed
5. checked out head from FreeBSD SVN
6. applied src/cddl diff
==> this failed as well

I have imported all nfse patch files from sourceforge in a local
mercurial repo to be able to easier follow what is changing. There I see
that cddl.diff was updated for the last time on May 17.

Could you help me with the following questions:
- Is INSTALL-all still relevant, and if yes, for which cases?
- What for is cddl.diff?
- I am heavily using zfs. Which patches from your patchset do I need to
get nfse to fully support zfs?

Lastly, I believe it might be more helpful to combine INSTALL-all and
INSTALL-kern into a single file INSTALL and in that file clearly point
out the differences between the two methods (what does one method give
you, what the other, what do I need to do for the first method, what for
the other).

Regards,

Martin

On 10/22/12 16:37, Andrey Simonenko wrote:
> On Sat, Oct 20, 2012 at 11:43:17AM +0200, Martin Birgmeier wrote:
>> Andrey,
>>
>> I'd really like to use this. However, I need to use it with FreeBSD 7.4,
>> 8.2, and 9.0 (and 9.1 in the near future); I tried to backport your
>> changes, but this turned out to be too difficult for me.
>>
> sys.diff for 8.2 in the attachment.

_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Mon, 29 Oct 2012 13:55:42 +0200

Hello,

On Fri, Oct 26, 2012 at 03:15:56PM +0200, Martin Birgmeier wrote:
> Hi Andrey,
>
> Today I started applying your changes and did the following:
>
> 1. downloaded nfse-20121025.tar.bz2 from sourceforge
> 2. read INSTALL-all
> 3. checked out release/8.2.0 from FreeBSD SVN
> 4. applied src/cddl.diff
> ==> this failed

The cddl.diff file cannot be applied to cddl/ from 8.2, since cddl/ source
was updated several times in next FreeBSD versions. These changes are not
strictly necessary (see below).

> 5. checked out head from FreeBSD SVN
> 6. applied src/cddl diff
> ==> this failed as well

Check again, I can apply cddl.diff to just csup'ed 10-CURRENT.

>
> I have imported all nfse patch files from sourceforge in a local
> mercurial repo to be able to easier follow what is changing. There I see
> that cddl.diff was updated for the last time on May 17.
>
> Could you help me with the following questions:
> - Is INSTALL-all still relevant, and if yes, for which cases?

This file describes how to apply all NFSE changes to the FreeBSD source
code to make complete integration.

> - What for is cddl.diff?
> - I am heavily using zfs. Which patches from your patchset do I need to
> get nfse to fully support zfs?

This file contains integration of NFSE with the zfs program. When one
calls 'zfs sharenfs/unshare ...', then NFS exports settings are updated
using SIGHUP or dynamic NFSE commands (depends on presence of the
/etc/nfs.exports file). If you do not use NFSE dynamic commands
(eg. "nfse -c 'flush/clear/add/update/delete/set/unset ...'"), then
cddl.diff is not needed, just create symlink to mountd.pid by setting
the nfse_mountd_pid rc variable to "YES" and 'zfs share/unshare ...'
will send SIGHUP to nfse.

There is one question about NFSE and ZFS, this is support of ZFS snapshots.
NFSE was implemented as part of NFS server, not the part of VFS framework.
As a result right now it is impossible to automatically (unconditionally,
as it is done in all FreeBSD versions) export ZFS snapshots by NFSE.

>
> Lastly, I believe it might be more helpful to combine INSTALL-all and
> INSTALL-kern into a single file INSTALL and in that file clearly point
> out the differences between the two methods (what does one method give
> you, what the other, what do I need to do for the first method, what for
> the other).

You are right, I've just updated these two files with descriptions.

I suggest to apply changes to sys/ (I sent to you before) and etc/
(necessary to correct two rejected updates). Then build and install
the kernel and try to run nfse with simple configuration (/etc/exports
or /etc/nfs.exports).
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)
The following reply was made to PR kern/136865; it has been noted by GNATS.

From: Andrey Simonenko <.ua>
To: Martin Birgmeier <>
Cc: bug-
Subject: Re: kern/136865: [nfs] [patch] NFS exports atomic and on-the-fly
atomic updates
Date: Tue, 30 Oct 2012 12:07:57 +0200

On Fri, Oct 26, 2012 at 03:15:56PM +0200, Martin Birgmeier wrote:

> - What for is cddl.diff?
> - I am heavily using zfs. Which patches from your patchset do I need to
> get nfse to fully support zfs?

I forgot to say, that cddl.diff also updates 'zfs sharenfs/unshare ...'
and zfs does not verifies correctness of exports(5) or nfs.exports(5)
settings. There are several PRs related to the current code in zfs
that verifies exports(5) settings.

If it is not necessary to use 'zfs sharenfs/unshare ...', and it is not
necessary to use NFSE dynamic commands (that are flushed after reloading
of export settings) and it is not necessary to use nfs.exports(5) file
format, then cddl.diff can be ignored.

The cddl.diff file has only changes to the zfs program and does not have
any changes related to ZFS in the kernel.
_______________________________________________
freebsd- mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-"
)