Hi all,
After merging the vfs-brauner tree, today's linux-next build (powerpc
allnoconfig) failed like this:
In file included from <command-line>:
In function 'i_size_read',
inlined from '__iomap_dio_rw' at fs/iomap/direct-io.c:570:16:
include/linux/compiler_types.h:435:45: error: call to '__compiletime_assert_229' declared with attribute error: Need native word sized stores/loads for atomicity.
435 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^
include/linux/compiler_types.h:416:25: note: in definition of macro '__compiletime_assert'
416 | prefix ## suffix(); \
| ^~~~~~
include/linux/compiler_types.h:435:9: note: in expansion of macro '_compiletime_assert'
435 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
include/linux/compiler_types.h:438:9: note: in expansion of macro 'compiletime_assert'
438 | compiletime_assert(__native_word(t), \
| ^~~~~~~~~~~~~~~~~~
include/asm-generic/barrier.h:206:9: note: in expansion of macro 'compiletime_assert_atomic_type'
206 | compiletime_assert_atomic_type(*p); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/linux/fs.h:911:16: note: in expansion of macro 'smp_load_acquire'
911 | return smp_load_acquire(&inode->i_size);
| ^~~~~~~~~~~~~~~~
Caused by commit
4bbd51d0f0ad ("fs: make the i_size_read/write helpers be smp_load_acquire/store_release()")
I have used the vfs-brauner tree from next-20240122 for today.
--
Cheers,
Stephen Rothwell
Hi all,
On Tue, 23 Jan 2024 12:52:27 +1100 Stephen Rothwell <[email protected]> wrote:
>
> After merging the vfs-brauner tree, today's linux-next build (powerpc
> allnoconfig) failed like this:
>
> In file included from <command-line>:
> In function 'i_size_read',
> inlined from '__iomap_dio_rw' at fs/iomap/direct-io.c:570:16:
> include/linux/compiler_types.h:435:45: error: call to '__compiletime_assert_229' declared with attribute error: Need native word sized stores/loads for atomicity.
> 435 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
> | ^
> include/linux/compiler_types.h:416:25: note: in definition of macro '__compiletime_assert'
> 416 | prefix ## suffix(); \
> | ^~~~~~
> include/linux/compiler_types.h:435:9: note: in expansion of macro '_compiletime_assert'
> 435 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
> | ^~~~~~~~~~~~~~~~~~~
> include/linux/compiler_types.h:438:9: note: in expansion of macro 'compiletime_assert'
> 438 | compiletime_assert(__native_word(t), \
> | ^~~~~~~~~~~~~~~~~~
> include/asm-generic/barrier.h:206:9: note: in expansion of macro 'compiletime_assert_atomic_type'
> 206 | compiletime_assert_atomic_type(*p); \
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> include/linux/fs.h:911:16: note: in expansion of macro 'smp_load_acquire'
> 911 | return smp_load_acquire(&inode->i_size);
> | ^~~~~~~~~~~~~~~~
>
> Caused by commit
>
> 4bbd51d0f0ad ("fs: make the i_size_read/write helpers be smp_load_acquire/store_release()")
>
> I have used the vfs-brauner tree from next-20240122 for today.
Pending a better resolution, today I have reverted that commit and the
following one.
--
Cheers,
Stephen Rothwell
On Wed, Jan 24, 2024 at 12:20:40PM +1100, Stephen Rothwell wrote:
> Hi all,
>
> On Tue, 23 Jan 2024 12:52:27 +1100 Stephen Rothwell <[email protected]> wrote:
> >
> > After merging the vfs-brauner tree, today's linux-next build (powerpc
> > allnoconfig) failed like this:
> >
> > In file included from <command-line>:
> > In function 'i_size_read',
> > inlined from '__iomap_dio_rw' at fs/iomap/direct-io.c:570:16:
> > include/linux/compiler_types.h:435:45: error: call to '__compiletime_assert_229' declared with attribute error: Need native word sized stores/loads for atomicity.
> > 435 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
> > | ^
> > include/linux/compiler_types.h:416:25: note: in definition of macro '__compiletime_assert'
> > 416 | prefix ## suffix(); \
> > | ^~~~~~
> > include/linux/compiler_types.h:435:9: note: in expansion of macro '_compiletime_assert'
> > 435 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
> > | ^~~~~~~~~~~~~~~~~~~
> > include/linux/compiler_types.h:438:9: note: in expansion of macro 'compiletime_assert'
> > 438 | compiletime_assert(__native_word(t), \
> > | ^~~~~~~~~~~~~~~~~~
> > include/asm-generic/barrier.h:206:9: note: in expansion of macro 'compiletime_assert_atomic_type'
> > 206 | compiletime_assert_atomic_type(*p); \
> > | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > include/linux/fs.h:911:16: note: in expansion of macro 'smp_load_acquire'
> > 911 | return smp_load_acquire(&inode->i_size);
> > | ^~~~~~~~~~~~~~~~
> >
> > Caused by commit
> >
> > 4bbd51d0f0ad ("fs: make the i_size_read/write helpers be smp_load_acquire/store_release()")
> >
> > I have used the vfs-brauner tree from next-20240122 for today.
>
> Pending a better resolution, today I have reverted that commit and the
> following one.
I had dropped both from vfs.misc yesterday night. Maybe it didn't make
it in time.
Hi Christian,
On Wed, 24 Jan 2024 12:13:02 +0100 Christian Brauner <[email protected]> wrote:
>
> I had dropped both from vfs.misc yesterday night. Maybe it didn't make
> it in time.
It still has not ... I only fetch your vfs.all branch. Did you remerge
and push it out?
--
Cheers,
Stephen Rothwell
On Wed, Jan 24, 2024 at 10:35:37PM +1100, Stephen Rothwell wrote:
> Hi Christian,
>
> On Wed, 24 Jan 2024 12:13:02 +0100 Christian Brauner <[email protected]> wrote:
> >
> > I had dropped both from vfs.misc yesterday night. Maybe it didn't make
> > it in time.
>
> It still has not ... I only fetch your vfs.all branch. Did you remerge
> and push it out?
So vfs.all should be at:
commit 8577a331532bb1d75f3536461739a0a8e15b219c
Merge: 0c5c260545bd 06b8db3a7dde
Author: Christian Brauner <[email protected]>
AuthorDate: Wed Jan 24 18:33:30 2024 +0100
Commit: Christian Brauner <[email protected]>
CommitDate: Wed Jan 24 18:33:30 2024 +0100
Merge branch 'vfs.fs' into vfs.all
Signed-off-by: Christian Brauner <[email protected]>
And that should've contain a merge of vfs.misc with all problematic
patches dropped. Please yell if you still see an issue tomorrow!