| Message ID | F19688880B763E40B28B2B462677FBF805BF0EAC8A@MX09A.corp.emc.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Note that the object layout also effectively needs the block layer. -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> -----Original Message----- > From: Christoph Hellwig [mailto:hch@infradead.org] > Sent: Thursday, August 11, 2011 2:07 PM > To: Peng, Tao > Cc: torvalds@linux-foundation.org; rdunlap@xenotime.net; bergwolf@gmail.com; > Trond.Myklebust@netapp.com; linux-raid@vger.kernel.org; > linux-kernel@vger.kernel.org; linux-nfs@vger.kernel.org > Subject: Re: Linux 3.1-rc1 (nfs/pnfs and drivers/md) > > Note that the object layout also effectively needs the block layer. > That is taken care of by PNFS_OBJLAYOUT depending on SCSI_OSD_ULD. The dependency chain for PNFS_OBJLAYOUT on BLOCK is: PNFS_OBJLAYOUT-> SCSI_OSD_ULD -> SCSI_OSD_INITIATOR -> SCSI-> BLOCK With the patch, if CONFIG_BLOCK is not set, pnfs file layout can still be built. Thanks, Tao -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 2011-08-11 06:26, tao.peng@emc.com wrote: > Hi, Randy and Linus > >> -----Original Message----- >> From: linux-nfs-owner@vger.kernel.org [mailto:linux-nfs-owner@vger.kernel.org] >> On Behalf Of Linus Torvalds >> Sent: Thursday, August 11, 2011 9:10 AM >> To: Randy Dunlap >> Cc: Peng Tao; Trond Myklebust; linux-raid@vger.kernel.org; Linux Kernel Mailing List; >> linux-nfs@vger.kernel.org >> Subject: Re: Linux 3.1-rc1 (nfs/pnfs and drivers/md) >> >> On Wed, Aug 10, 2011 at 5:58 PM, Randy Dunlap <rdunlap@xenotime.net> wrote: >>> >>> The blocklayout parts of NVS_V4_1 could depend on BLOCK. or all of NVS_V4_1 >>> can depend on BLOCK. >> >> Absolutely. Replace "could" by "should". >> >> There's no way I'll take a patch to make NFS "select BLOCK". If the >> block layer isn't enabled, then NFS obviously shouldn't be doing any >> pnfs stuff. > I see. How about the following patch? We used to do like following for PNFS_BLOCK but it was changed during code review. My Ack, that's indeed closer to what I had queued in my tree originally and follows the approach we used for the objects layout driver. Benny > > > From 833b59c1d24b43a9a4d47f777abab2cd9d041c68 Mon Sep 17 00:00:00 2001 > From: Peng Tao <bergwolf@gmail.com> > Date: Wed, 10 Aug 2011 18:29:21 -0400 > Subject: [PATCH] NFS41: make PNFS_BLOCK selectable > > PNFS_BLOCK needs BLK_DEV_DM/MD, which is not a dependency for other > pnfs layout drivers. Seperate it out so others can still build when > BLK_DEV_DM/MD is not enabled. > > Also change select to depends on to avoid build failures. > > Reported-by: Randy Dunlap <rdunlap@xenotime.net> > Signed-off-by: Peng Tao <peng_tao@emc.com> > --- > fs/nfs/Kconfig | 14 +++++++++----- > 1 files changed, 9 insertions(+), 5 deletions(-) > > diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig > index be02077..0d30613 100644 > --- a/fs/nfs/Kconfig > +++ b/fs/nfs/Kconfig > @@ -79,12 +79,9 @@ config NFS_V4_1 > depends on NFS_FS && NFS_V4 && EXPERIMENTAL > select SUNRPC_BACKCHANNEL > select PNFS_FILE_LAYOUT > - select PNFS_BLOCK > - select MD > - select BLK_DEV_DM > help > This option enables support for minor version 1 of the NFSv4 protocol > - (RFC 5661 and RFC 5663) in the kernel's NFS client. > + (RFC 5661) in the kernel's NFS client. > > If unsure, say N. > > @@ -92,7 +89,14 @@ config PNFS_FILE_LAYOUT > tristate > > config PNFS_BLOCK > - tristate > + tristate "Provide support for the pNFS Block Layout Driver for NFSv4.1 pNFS (EXPERIMENTAL)" > + depends on NFS_FS && NFS_V4_1 && BLK_DEV_DM > + help > + Say M here if you want your pNFS client to support the Block Layout Driver > + (RFC 5663). Requires Multiple devices driver support (DM) and Device mapper > + support (BLK_DEV_DM). > + > + If unsure, say N. > > config PNFS_OBJLAYOUT > tristate "Provide support for the pNFS Objects Layout Driver for NFSv4.1 pNFS (EXPERIMENTAL)" -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, 10 Aug 2011 23:26:47 -0400 tao.peng@emc.com wrote: > Hi, Randy and Linus > > > -----Original Message----- > > From: linux-nfs-owner@vger.kernel.org [mailto:linux-nfs-owner@vger.kernel.org] > > On Behalf Of Linus Torvalds > > Sent: Thursday, August 11, 2011 9:10 AM > > To: Randy Dunlap > > Cc: Peng Tao; Trond Myklebust; linux-raid@vger.kernel.org; Linux Kernel Mailing List; > > linux-nfs@vger.kernel.org > > Subject: Re: Linux 3.1-rc1 (nfs/pnfs and drivers/md) > > > > On Wed, Aug 10, 2011 at 5:58 PM, Randy Dunlap <rdunlap@xenotime.net> wrote: > > > > > > The blocklayout parts of NVS_V4_1 could depend on BLOCK. or all of NVS_V4_1 > > > can depend on BLOCK. > > > > Absolutely. Replace "could" by "should". > > > > There's no way I'll take a patch to make NFS "select BLOCK". If the > > block layer isn't enabled, then NFS obviously shouldn't be doing any > > pnfs stuff. > I see. How about the following patch? We used to do like following for PNFS_BLOCK but it was changed during code review. Yes, that builds cleanly for me. Thanks. Acked-by: Randy Dunlap <rdunlap@xenotime.net> > From 833b59c1d24b43a9a4d47f777abab2cd9d041c68 Mon Sep 17 00:00:00 2001 > From: Peng Tao <bergwolf@gmail.com> > Date: Wed, 10 Aug 2011 18:29:21 -0400 > Subject: [PATCH] NFS41: make PNFS_BLOCK selectable > > PNFS_BLOCK needs BLK_DEV_DM/MD, which is not a dependency for other > pnfs layout drivers. Seperate it out so others can still build when > BLK_DEV_DM/MD is not enabled. > > Also change select to depends on to avoid build failures. > > Reported-by: Randy Dunlap <rdunlap@xenotime.net> > Signed-off-by: Peng Tao <peng_tao@emc.com> > --- > fs/nfs/Kconfig | 14 +++++++++----- > 1 files changed, 9 insertions(+), 5 deletions(-) > > diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig > index be02077..0d30613 100644 > --- a/fs/nfs/Kconfig > +++ b/fs/nfs/Kconfig > @@ -79,12 +79,9 @@ config NFS_V4_1 > depends on NFS_FS && NFS_V4 && EXPERIMENTAL > select SUNRPC_BACKCHANNEL > select PNFS_FILE_LAYOUT > - select PNFS_BLOCK > - select MD > - select BLK_DEV_DM > help > This option enables support for minor version 1 of the NFSv4 protocol > - (RFC 5661 and RFC 5663) in the kernel's NFS client. > + (RFC 5661) in the kernel's NFS client. > > If unsure, say N. > > @@ -92,7 +89,14 @@ config PNFS_FILE_LAYOUT > tristate > > config PNFS_BLOCK > - tristate > + tristate "Provide support for the pNFS Block Layout Driver for NFSv4.1 pNFS (EXPERIMENTAL)" > + depends on NFS_FS && NFS_V4_1 && BLK_DEV_DM > + help > + Say M here if you want your pNFS client to support the Block Layout Driver > + (RFC 5663). Requires Multiple devices driver support (DM) and Device mapper > + support (BLK_DEV_DM). > + > + If unsure, say N. > > config PNFS_OBJLAYOUT > tristate "Provide support for the pNFS Objects Layout Driver for NFSv4.1 pNFS (EXPERIMENTAL)" > -- > 1.7.4.2 > > > --- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 08/11/2011 03:11 AM, Benny Halevy wrote: > On 2011-08-11 06:26, tao.peng@emc.com wrote: > > My Ack, that's indeed closer to what I had queued in my tree originally and > follows the approach we used for the objects layout driver. > > Benny > Yes but this breaks "Sorin"'s conspiracy to sneak into readhat the block layout driver. Because it would be selected automatically if you set CONFIG_NFS41. Because for some perverted reason readhat wants pNFS-Files in, with latest patches, but not the blocks or objects. I liked it that Blocks tried to state: "pNFS is with blocks as well". Even Trond agreed that it is logical this way. You select CONFIG_NFS41 or not. objects and blocks are a part of that, mandated by the protocol. (I don't know that it was "Sorin"'s Idea, I meant the block guys) I wish we can go back to the auto selection but fix the dependency stuff. Surly it is possible. Thanks Boaz -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index be02077..0d30613 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -79,12 +79,9 @@ config NFS_V4_1 depends on NFS_FS && NFS_V4 && EXPERIMENTAL select SUNRPC_BACKCHANNEL select PNFS_FILE_LAYOUT - select PNFS_BLOCK - select MD - select BLK_DEV_DM help This option enables support for minor version 1 of the NFSv4 protocol - (RFC 5661 and RFC 5663) in the kernel's NFS client. + (RFC 5661) in the kernel's NFS client. If unsure, say N. @@ -92,7 +89,14 @@ config PNFS_FILE_LAYOUT tristate config PNFS_BLOCK - tristate + tristate "Provide support for the pNFS Block Layout Driver for NFSv4.1 pNFS (EXPERIMENTAL)" + depends on NFS_FS && NFS_V4_1 && BLK_DEV_DM + help + Say M here if you want your pNFS client to support the Block Layout Driver + (RFC 5663). Requires Multiple devices driver support (DM) and Device mapper + support (BLK_DEV_DM). + + If unsure, say N. config PNFS_OBJLAYOUT tristate "Provide support for the pNFS Objects Layout Driver for NFSv4.1 pNFS (EXPERIMENTAL)"