[PATCH 4.9] dm bufio: avoid false-positive Wmaybe-uninitialized warning

[PATCH 4.9] dm bufio: avoid false-positive Wmaybe-uninitialized warning

[Nathan Chancellor]

From: Arnd Bergmann <arnd@arndb.de>

commit 590347e4000356f55eb10b03ced2686bd74dab40 upstream.

gcc-6.3 and earlier show a new warning after a seemingly unrelated
change to the arm64 PAGE_KERNEL definition:

In file included from drivers/md/dm-bufio.c:14:0:
drivers/md/dm-bufio.c: In function 'alloc_buffer':
include/linux/sched/mm.h:182:56: warning: 'noio_flag' may be used uninitialized in this function [-Wmaybe-uninitialized]
  current->flags = (current->flags & ~PF_MEMALLOC_NOIO) | flags;
                                                        ^

The same warning happened earlier on linux-3.18 for MIPS and I did a
workaround for that, but now it's come back.

gcc-7 and newer are apparently smart enough to figure this out, and
other architectures don't show it, so the best I could come up with is
to rework the caller slightly in a way that makes it obvious enough to
all arm64 compilers what is happening here.

Fixes: 41acec624087 ("arm64: kpti: Make use of nG dependent on arm64_kernel_unmapped_at_el0()")
Link: https://patchwork.kernel.org/patch/9692829/
Cc: stable@vger.kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
[snitzer: moved declarations inside conditional, altered vmalloc return]
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
[nc: Backport to 4.9, adjust context for lack of 19809c2da28a]
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
---

Hi Greg,

Resending this with a proper prefix and message, in case the other one
fell through the cracks. I came across this warning when building with
Google's stock GCC 4.9 toolchain on the OnePlus 6. Context was adjusted
around lack of commit 19809c2da28a ("mm, vmalloc: use __GFP_HIGHMEM
implicitly") in 4.9. Please apply when you get a chance.

Thanks!
Nathan

 drivers/md/dm-bufio.c | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

diff --git a/drivers/md/dm-bufio.c b/drivers/md/dm-bufio.c
index 3ec647e8b9c6..35fd57fdeba9 100644
--- a/drivers/md/dm-bufio.c
+++ b/drivers/md/dm-bufio.c
@@ -373,9 +373,6 @@ static void __cache_size_refresh(void)
 static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask,
 			       enum data_mode *data_mode)
 {
-	unsigned noio_flag;
-	void *ptr;
-
 	if (c->block_size <= DM_BUFIO_BLOCK_SIZE_SLAB_LIMIT) {
 		*data_mode = DATA_MODE_SLAB;
 		return kmem_cache_alloc(DM_BUFIO_CACHE(c), gfp_mask);
@@ -399,16 +396,16 @@ static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask,
 	 * all allocations done by this process (including pagetables) are done
 	 * as if GFP_NOIO was specified.
 	 */
+	if (gfp_mask & __GFP_NORETRY) {
+		unsigned noio_flag = memalloc_noio_save();
+		void *ptr = __vmalloc(c->block_size, gfp_mask | __GFP_HIGHMEM,
+				      PAGE_KERNEL);
 
-	if (gfp_mask & __GFP_NORETRY)
-		noio_flag = memalloc_noio_save();
-
-	ptr = __vmalloc(c->block_size, gfp_mask | __GFP_HIGHMEM, PAGE_KERNEL);
-
-	if (gfp_mask & __GFP_NORETRY)
 		memalloc_noio_restore(noio_flag);
+		return ptr;
+	}
 
-	return ptr;
+	return __vmalloc(c->block_size, gfp_mask | __GFP_HIGHMEM, PAGE_KERNEL);
 }
 
 /*
-- 
2.17.1

Re: [PATCH 4.9] dm bufio: avoid false-positive Wmaybe-uninitialized warning

[Greg Kroah-Hartman]

On Mon, Jun 11, 2018 at 09:05:33AM -0700, Nathan Chancellor wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> commit 590347e4000356f55eb10b03ced2686bd74dab40 upstream.
> 
> gcc-6.3 and earlier show a new warning after a seemingly unrelated
> change to the arm64 PAGE_KERNEL definition:
> 
> In file included from drivers/md/dm-bufio.c:14:0:
> drivers/md/dm-bufio.c: In function 'alloc_buffer':
> include/linux/sched/mm.h:182:56: warning: 'noio_flag' may be used uninitialized in this function [-Wmaybe-uninitialized]
>   current->flags = (current->flags & ~PF_MEMALLOC_NOIO) | flags;
>                                                         ^
> 
> The same warning happened earlier on linux-3.18 for MIPS and I did a
> workaround for that, but now it's come back.
> 
> gcc-7 and newer are apparently smart enough to figure this out, and
> other architectures don't show it, so the best I could come up with is
> to rework the caller slightly in a way that makes it obvious enough to
> all arm64 compilers what is happening here.
> 
> Fixes: 41acec624087 ("arm64: kpti: Make use of nG dependent on arm64_kernel_unmapped_at_el0()")
> Link: https://patchwork.kernel.org/patch/9692829/
> Cc: stable@vger.kernel.org
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> [snitzer: moved declarations inside conditional, altered vmalloc return]
> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
> [nc: Backport to 4.9, adjust context for lack of 19809c2da28a]
> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
> ---
> 
> Hi Greg,
> 
> Resending this with a proper prefix and message, in case the other one
> fell through the cracks. I came across this warning when building with
> Google's stock GCC 4.9 toolchain on the OnePlus 6. Context was adjusted
> around lack of commit 19809c2da28a ("mm, vmalloc: use __GFP_HIGHMEM
> implicitly") in 4.9. Please apply when you get a chance.

Now queued up, thanks.  The other one was still in my queue, give me a
chance to catch up :)

greg k-h

Re: [PATCH 4.9] dm bufio: avoid false-positive Wmaybe-uninitialized warning

[Nathan Chancellor]

On Tue, Jun 12, 2018 at 06:12:13PM +0200, Greg Kroah-Hartman wrote:
> On Mon, Jun 11, 2018 at 09:05:33AM -0700, Nathan Chancellor wrote:
> > From: Arnd Bergmann <arnd@arndb.de>
> > 
> > commit 590347e4000356f55eb10b03ced2686bd74dab40 upstream.
> > 
> > gcc-6.3 and earlier show a new warning after a seemingly unrelated
> > change to the arm64 PAGE_KERNEL definition:
> > 
> > In file included from drivers/md/dm-bufio.c:14:0:
> > drivers/md/dm-bufio.c: In function 'alloc_buffer':
> > include/linux/sched/mm.h:182:56: warning: 'noio_flag' may be used uninitialized in this function [-Wmaybe-uninitialized]
> >   current->flags = (current->flags & ~PF_MEMALLOC_NOIO) | flags;
> >                                                         ^
> > 
> > The same warning happened earlier on linux-3.18 for MIPS and I did a
> > workaround for that, but now it's come back.
> > 
> > gcc-7 and newer are apparently smart enough to figure this out, and
> > other architectures don't show it, so the best I could come up with is
> > to rework the caller slightly in a way that makes it obvious enough to
> > all arm64 compilers what is happening here.
> > 
> > Fixes: 41acec624087 ("arm64: kpti: Make use of nG dependent on arm64_kernel_unmapped_at_el0()")
> > Link: https://patchwork.kernel.org/patch/9692829/
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> > [snitzer: moved declarations inside conditional, altered vmalloc return]
> > Signed-off-by: Mike Snitzer <snitzer@redhat.com>
> > [nc: Backport to 4.9, adjust context for lack of 19809c2da28a]
> > Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
> > ---
> > 
> > Hi Greg,
> > 
> > Resending this with a proper prefix and message, in case the other one
> > fell through the cracks. I came across this warning when building with
> > Google's stock GCC 4.9 toolchain on the OnePlus 6. Context was adjusted
> > around lack of commit 19809c2da28a ("mm, vmalloc: use __GFP_HIGHMEM
> > implicitly") in 4.9. Please apply when you get a chance.
> 
> Now queued up, thanks.  The other one was still in my queue, give me a
> chance to catch up :)
> 
> greg k-h

Sorry, didn't realize you were behind :/

Thanks for the timely response on this one!
Nathan