lib: Move find_last_bit.o to obj-y to enable use by modules.
Subsystems (2)
PATCH 1/1 lib: Move find_last_bit.o to obj-y to enable use by modules.
Caught with the sh allmodconfig:
ERROR: "find_last_bit" [fs/nfs/nfs.ko] undefined! make[2]: *** [__modpost] Error 1 make[1]: *** [modules] Error 2 make: *** [sub-make] Error 2
find_last_bit.o is currently built with lib-y, which ends up breaking the nfs module build after the ("nfs41: free slot") commit. Move it to obj-y so the EXPORT_SYMBOL() actually has some effect.
---
lib/Makefile +1 -1
@@ -38,7 +38,7 @@ lib-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o lib-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem.o lib-$(CONFIG_GENERIC_FIND_FIRST_BIT) += find_next_bit.o lib-$(CONFIG_GENERIC_FIND_NEXT_BIT) += find_next_bit.o -lib-$(CONFIG_GENERIC_FIND_LAST_BIT) += find_last_bit.o +obj-$(CONFIG_GENERIC_FIND_LAST_BIT) += find_last_bit.o obj-$(CONFIG_GENERIC_HWEIGHT) += hweight.o obj-$(CONFIG_LOCK_KERNEL) += kernel_lock.o obj-$(CONFIG_DEBUG_PREEMPT) += smp_processor_id.o
Quoted 7 lines · Paul Mundt
Caught with the sh allmodconfig:
ERROR: "find_last_bit" [fs/nfs/nfs.ko] undefined! make[2]: *** [__modpost] Error 1 make[1]: *** [modules] Error 2 make: *** [sub-make] Error 2
Quoted 6 lines
find_last_bit.o is currently built with lib-y, which ends up breaking the nfs module build after the ("nfs41: free slot") commit. Move it to obj-y so the
EXPORT_SYMBOL()actually has some effect.Signed-off-by Paul Mundt <[email protected]>Cc Benny Halevy <[email protected]>
ACK. and thanks!
FYI, Fred's original patch can be found here:
http://patchwork.kernel.org/patch/14572/
It is also queued in the linux-pnfs tree:
http://git.linux-nfs.org/?p=bhalevy/linux-pnfs.git;a=commitdiff;h=1e3a7552d9de2ba101b76deed99605f0145fc4d5
but I haven't submitted it to Trond since I expected it to get upstream (and to -next) via linux-kbuild.
Trond, would you like to pull this change to your nfsv41 branch? (should appear before "nfs41: free slot" as Paul noticed)
Benny
Quoted 18 lines
Cc Andy Adamson <[email protected]>Cc Trond Myklebust <[email protected]>---
lib/Makefile +1
-1@@ -38,7 +38,7 @@ lib-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o lib-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem.o lib-$(CONFIG_GENERIC_FIND_FIRST_BIT) += find_next_bit.o lib-$(CONFIG_GENERIC_FIND_NEXT_BIT) += find_next_bit.o -lib-$(CONFIG_GENERIC_FIND_LAST_BIT) += find_last_bit.o +obj-$(CONFIG_GENERIC_FIND_LAST_BIT) += find_last_bit.o obj-$(CONFIG_GENERIC_HWEIGHT) += hweight.o obj-$(CONFIG_LOCK_KERNEL) += kernel_lock.o obj-$(CONFIG_DEBUG_PREEMPT) += smp_processor_id.o
Quoted 20 lines · Benny Halevy
Quoted 7 lines · Paul Mundt
Caught with the sh allmodconfig:
ERROR: "find_last_bit" [fs/nfs/nfs.ko] undefined! make[2]: *** [__modpost] Error 1 make[1]: *** [modules] Error 2 make: *** [sub-make] Error 2Quoted 6 lines
find_last_bit.o is currently built with lib-y, which ends up breaking the nfs module build after the ("nfs41: free slot") commit. Move it to obj-y so the
EXPORT_SYMBOL()actually has some effect.Signed-off-by Paul Mundt <[email protected]>Cc Benny Halevy <[email protected]>ACK. and thanks!
FYI, Fred's original patch can be found here:
http://patchwork.kernel.org/patch/14572/
Quoted 8 lines
It is also queued in the linux-pnfs tree:
http://git.linux-nfs.org/?p=bhalevy/linux-pnfs.git;a=commitdiff;h=1e3a7552d9de2ba101b76deed99605f0145fc4d5
but I haven't submitted it to Trond since I expected it to get upstream (and to -next) via linux-kbuild.Trond, would you like to pull this change to your nfsv41 branch? (should appear before "nfs41: free slot" as Paul noticed)
Ok, so we have two different trivial patches for fixing the same thing, and a week later it is still broken.
I realize it is a trivial patch, but it does break builds. If folks aren't going to take these sorts of things more seriously, then their tree should be dropped after a grace period (say 2 days or so).
Beyond that, it doesn't seem like -next has any sort of coherent policy for dealing with trivial patches. If the emphasis is on the tree that introduced the regression to deal with it, then trees need to be aggressively dropped when these things go unfixed.
Having builds broken for a week for an issue that has been spotted and fixed by several people is simply unacceptable.
Quoted 35 lines · Paul Mundt
Quoted 19 lines · Benny Halevy
Quoted 7 lines · Paul Mundt
Caught with the sh allmodconfig:
ERROR: "find_last_bit" [fs/nfs/nfs.ko] undefined! make[2]: *** [__modpost] Error 1 make[1]: *** [modules] Error 2 make: *** [sub-make] Error 2Quoted 6 lines
find_last_bit.o is currently built with lib-y, which ends up breaking the nfs module build after the ("nfs41: free slot") commit. Move it to obj-y so the
EXPORT_SYMBOL()actually has some effect.Signed-off-by Paul Mundt <[email protected]>Cc Benny Halevy <[email protected]>ACK. and thanks!
FYI, Fred's original patch can be found here:
http://patchwork.kernel.org/patch/14572/Quoted 8 lines
It is also queued in the linux-pnfs tree:
http://git.linux-nfs.org/?p=bhalevy/linux-pnfs.git;a=commitdiff;h=1e3a7552d9de2ba101b76deed99605f0145fc4d5
but I haven't submitted it to Trond since I expected it to get upstream (and to -next) via linux-kbuild.Trond, would you like to pull this change to your nfsv41 branch? (should appear before "nfs41: free slot" as Paul noticed)
Ok, so we have two different trivial patches for fixing the same thing, and a week later it is still broken.
I realize it is a trivial patch, but it does break builds. If folks aren't going to take these sorts of things more seriously, then their tree should be dropped after a grace period (say 2 days or so).
Quoted 7 lines
Beyond that, it doesn't seem like -next has any sort of coherent policy for dealing with trivial patches. If the emphasis is on the tree that introduced the regression to deal with it, then trees need to be aggressively dropped when these things go unfixed.
Having builds broken for a week for an issue that has been spotted and fixed by several people is simply unacceptable.
Paul, that's a valid point but I don't set these polices. Trond suggested to just commit this to 2.6.30 and I asked Rusty's Ack here:
http://lkml.org/lkml/2009/4/21/489
Like I said there, I'm not sure who to send this patch to. Ingo?
Benny
Quoted 16 lines · Benny Halevy
Quoted 7 lines · Paul Mundt
Ok, so we have two different trivial patches for fixing the same thing, and a week later it is still broken.
I realize it is a trivial patch, but it does break builds. If folks aren't going to take these sorts of things more seriously, then their tree should be dropped after a grace period (say 2 days or so).
Quoted 7 lines
Beyond that, it doesn't seem like -next has any sort of coherent policy for dealing with trivial patches. If the emphasis is on the tree that introduced the regression to deal with it, then trees need to be aggressively dropped when these things go unfixed.
Having builds broken for a week for an issue that has been spotted and fixed by several people is simply unacceptable.
Quoted 8 lines
Paul, that's a valid point but I don't set these polices. Trond suggested to just commit this to 2.6.30 and I asked Rusty's Ack here:
http://lkml.org/lkml/2009/4/21/489Like I said there, I'm not sure who to send this patch to. Ingo?
I was under the impression that a tree that caused a build regression would be dropped until it had it sorted out, but that seems to be more the exception than the rule.
-next is good at finding bugs in build configurations folks haven't considered, which should serve as a pretty good platform for getting those types of fixes merged quickly, whether it be in to the tree that caused the regression or -next directly.
Unfortunately it seems like build regressions are more of an afterthought than a show stopper. I count at least 3 on the sh builds in the last couple weeks that are all averaging a week or longer to unbreak, while patches have been available almost immediately.
Quoted 34 lines · Paul Mundt
Quoted 16 lines · Benny Halevy
Quoted 7 lines · Paul Mundt
Ok, so we have two different trivial patches for fixing the same thing, and a week later it is still broken.
I realize it is a trivial patch, but it does break builds. If folks aren't going to take these sorts of things more seriously, then their tree should be dropped after a grace period (say 2 days or so).
Quoted 7 lines
Beyond that, it doesn't seem like -next has any sort of coherent policy for dealing with trivial patches. If the emphasis is on the tree that introduced the regression to deal with it, then trees need to be aggressively dropped when these things go unfixed.
Having builds broken for a week for an issue that has been spotted and fixed by several people is simply unacceptable.
Quoted 8 lines
Paul, that's a valid point but I don't set these polices. Trond suggested to just commit this to 2.6.30 and I asked Rusty's Ack here:
http://lkml.org/lkml/2009/4/21/489Like I said there, I'm not sure who to send this patch to. Ingo?
I was under the impression that a tree that caused a build regression would be dropped until it had it sorted out, but that seems to be more the exception than the rule.
-next is good at finding bugs in build configurations folks haven't considered, which should serve as a pretty good platform for getting those types of fixes merged quickly, whether it be in to the tree that caused the regression or -next directly.
Quoted 4 lines
Unfortunately it seems like build regressions are more of an afterthought than a show stopper. I count at least 3 on the sh builds in the last couple weeks that are all averaging a week or longer to unbreak, while patches have been available almost immediately.
In this case, the tree in question is exposing a bug that already exists in mainline; a function that is explicitly labelled as being exported for use by arbitrary modules, and yet isn't being compiled into the kernel. Shooting the messenger isn't going to fix that.
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
Trond
Signature · 6 lines
Trond Myklebust Linux NFS client maintainer NetApp [email protected] www.netapp.com
Quoted 39 lines · Trond Myklebust
Quoted 33 lines · Paul Mundt
Quoted 15 lines · Benny Halevy
Quoted 7 lines · Paul Mundt
Ok, so we have two different trivial patches for fixing the same thing, and a week later it is still broken.
I realize it is a trivial patch, but it does break builds. If folks aren't going to take these sorts of things more seriously, then their tree should be dropped after a grace period (say 2 days or so).
Quoted 7 lines
Beyond that, it doesn't seem like -next has any sort of coherent policy for dealing with trivial patches. If the emphasis is on the tree that introduced the regression to deal with it, then trees need to be aggressively dropped when these things go unfixed.
Having builds broken for a week for an issue that has been spotted and fixed by several people is simply unacceptable.
Quoted 8 lines
Paul, that's a valid point but I don't set these polices. Trond suggested to just commit this to 2.6.30 and I asked Rusty's Ack here:
http://lkml.org/lkml/2009/4/21/489Like I said there, I'm not sure who to send this patch to. Ingo?
I was under the impression that a tree that caused a build regression would be dropped until it had it sorted out, but that seems to be more the exception than the rule.
-next is good at finding bugs in build configurations folks haven't considered, which should serve as a pretty good platform for getting those types of fixes merged quickly, whether it be in to the tree that caused the regression or -next directly.
Quoted 4 lines
Unfortunately it seems like build regressions are more of an afterthought than a show stopper. I count at least 3 on the sh builds in the last couple weeks that are all averaging a week or longer to unbreak, while patches have been available almost immediately.
Quoted 8 lines
In this case, the tree in question is exposing a bug that already exists in mainline; a function that is explicitly labelled as being exported for use by arbitrary modules, and yet isn't being compiled into the kernel. Shooting the messenger isn't going to fix that.
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
Will do.
Regardless, Paul's observation seems valid.
I wonder if linux-next should have a branch or pull a tree holding trivial fixes fitting no other specific tree...
Benny
Quoted 2 lines
Trond
Quoted 15 lines · Trond Myklebust
Quoted 9 lines · Paul Mundt
I was under the impression that a tree that caused a build regression would be dropped until it had it sorted out, but that seems to be more the exception than the rule.
-next is good at finding bugs in build configurations folks haven't considered, which should serve as a pretty good platform for getting those types of fixes merged quickly, whether it be in to the tree that caused the regression or -next directly.
Quoted 4 lines
Unfortunately it seems like build regressions are more of an afterthought than a show stopper. I count at least 3 on the sh builds in the last couple weeks that are all averaging a week or longer to unbreak, while patches have been available almost immediately.
Quoted 5 lines
In this case, the tree in question is exposing a bug that already exists in mainline; a function that is explicitly labelled as being exported for use by arbitrary modules, and yet isn't being compiled into the kernel. Shooting the messenger isn't going to fix that.
Nor is bantering on semantics going to change the fact that you caused a build regression and have so far refused to even acknowledge that that is a problem. If anyone had bothered to test the module build, this would have shown up immediately and would have rightly been fixed before that change was merged. Claiming that the two are completely unrelated after the fact is rather disingenuous, and does nothing to address the process problem.
Quoted 4 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
That's a complete cop-out, if there had been no export at all how would you have proceeded? How is this situation any different?
I am willing to monitor my builds regularly and send out trivial patches to fix whatever ends up breaking, but I do expect people who have caused regressions to take these matters seriously rather than sweep them under the carpet and pretend like they either don't exist or are somehow someone else's problem.
Any tree that causes a regression should simply be dropped until it gets it sorted out, as it's abundantly obvious that people aren't even prepared to do the bare minimum to keep things working after they've been pulled from.
Quoted 6 lines · Paul Mundt
Quoted 4 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
That's a complete cop-out, if there had been no export at all how would you have proceeded? How is this situation any different?
Please see Documentation/SubmittingPatches. If we were requesting a new kernel feature, I'd go looking for the maintainer and get an ACK or get said maintainer to include the patch in their tree.
As it is, said kernel feature exists, but is broken. The correct thing to do is to fix the kernel feature. Sooner, rather than later.
Signature · 6 lines
Trond Myklebust Linux NFS client maintainer NetApp [email protected] www.netapp.com
Quoted 8 lines · Trond Myklebust
Quoted 6 lines · Paul Mundt
Quoted 4 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
That's a complete cop-out, if there had been no export at all how would you have proceeded? How is this situation any different?
Quoted 7 lines
Please see
Documentation/SubmittingPatches. If we were requesting a new kernel feature, I'd go looking for the maintainer and get an ACK or get said maintainer to include the patch in their tree.As it is, said kernel feature exists, but is broken. The correct thing to do is to fix the kernel feature. Sooner, rather than later.
And until that feature is working, you have absolutely no business merging crap in to the kernel that uses it. You made an assumption that it was working, which turned out to be wrong (which you also failed to validate through any form of testing), and subsequently caused a regression. Mistakes happen and are expected in -next, refusing to deal with them at all on the other hand..
The first course of action any responsible maintainer would have taken would be to revert the offending changeset once the issue was pointed out until its dependencies were sorted out. Not only have you failed to do that, you couldn't even be bothered replying to the email with patch that pointed out the build failure in the first place. This created the mistaken conclusion that you were actually doing something about the regression you caused rather than just dismissing it out of hand.
The nfs tree should be dropped on the floor immediately until this issue is resolved.
Quoted 31 lines · Paul Mundt
Quoted 8 lines · Trond Myklebust
Quoted 6 lines · Paul Mundt
Quoted 4 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
That's a complete cop-out, if there had been no export at all how would you have proceeded? How is this situation any different?
Quoted 7 lines
Please see
Documentation/SubmittingPatches. If we were requesting a new kernel feature, I'd go looking for the maintainer and get an ACK or get said maintainer to include the patch in their tree.As it is, said kernel feature exists, but is broken. The correct thing to do is to fix the kernel feature. Sooner, rather than later.
And until that feature is working, you have absolutely no business merging crap in to the kernel that uses it. You made an assumption that it was working, which turned out to be wrong (which you also failed to validate through any form of testing), and subsequently caused a regression. Mistakes happen and are expected in -next, refusing to deal with them at all on the other hand..
The first course of action any responsible maintainer would have taken would be to revert the offending changeset once the issue was pointed out until its dependencies were sorted out. Not only have you failed to do that, you couldn't even be bothered replying to the email with patch that pointed out the build failure in the first place. This created the mistaken conclusion that you were actually doing something about the regression you caused rather than just dismissing it out of hand.
Quoted 2 lines
The nfs tree should be dropped on the floor immediately until this issue is resolved.
Whatever...
Signature · 6 lines
Trond Myklebust Linux NFS client maintainer NetApp [email protected] www.netapp.com
Quoted 4 lines · Trond Myklebust
Quoted 4 lines
Quoted 4 lines · Paul Mundt
Quoted 3 lines
Quoted 3 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
...
Quoted 4 lines
Quoted 2 lines
The nfs tree should be dropped on the floor immediately until this issue is resolved.
Whatever...
FWIW Trond, I would have taken the patch and pushed it. It happens: it's not your fault, but it suddenly became your problem.
And Paul: being right doesn't excuse being an asshole.[1]
Thanks,
Rusty.
[1] As this rule applies to this message, I ponder my own failings.
Quoted 5 lines · Rusty Russell
Quoted 4 lines · Trond Myklebust
Quoted 4 lines
Quoted 4 lines · Paul Mundt
Quoted 3 lines
Quoted 3 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
Quoted 8 lines
...
Quoted 4 lines
Quoted 2 lines
The nfs tree should be dropped on the floor immediately until this issue is resolved.
Whatever...
FWIW Trond, I would have taken the patch and pushed it. It
happens: it's not your fault, but it suddenly became your problem.
To double check things: is it clear now who pushes the fix and when? I'm obviously acking it and can push it to Linus if needed.
Thanks,
Ingo
Quoted 11 lines · Ingo Molnar
Quoted 9 lines · Rusty Russell
Quoted 4 lines · Trond Myklebust
Quoted 4 lines
Quoted 4 lines · Paul Mundt
Quoted 3 lines
Quoted 3 lines
In any case, this patch does not belong in the NFS tree since it touches generic library code, not NFS code. Benny, if nobody else wants to shepherd it, then just send it directly to Linus.
...
Quoted 3 lines
Quoted 2 lines
The nfs tree should be dropped on the floor immediately until this issue is resolved.
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
Whatever...
FWIW Trond, I would have taken the patch and pushed it. It
happens: it's not your fault, but it suddenly became your problem.To double check things: is it clear now who pushes the fix and when? I'm obviously acking it and can push it to Linus if needed.
Linus has already merged it as commita5422a5111811401f7756345e4c237ff06cf6d1e
Cheers
Trond
Signature · 6 lines
Trond Myklebust Linux NFS client maintainer NetApp [email protected] www.netapp.com