[PATCH][next] ext4: avoid dozens of -Wflex-array-member-not-at-end warnings

[PATCH][next] ext4: avoid dozens of -Wflex-array-member-not-at-end warnings

[Gustavo A. R. Silva]

-Wflex-array-member-not-at-end was introduced in GCC-14, and we
are getting ready to enable it, globally.

Use the `DEFINE_RAW_FLEX()` helper for an on-stack definition of
a flexible structure (`struct shash_desc`) where the size of the
flexible-array member (`__ctx`) is known at compile-time, and
refactor the rest of the code, accordingly.

So, with these changes, fix 38 of the following warnings:

fs/ext4/ext4.h:2471:35: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
---

Worth mentioning is that this issue is pretty much the same as this
other one:

	https://git.kernel.org/linus/2bd9077b6261

Thanks!

 fs/ext4/ext4.h | 20 +++++++++++---------
 1 file changed, 11 insertions(+), 9 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 74f2071189b2..edcbe189ff34 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2460,22 +2460,24 @@ static inline __le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize)
 #define DX_HASH_SIPHASH			6
 #define DX_HASH_LAST 			DX_HASH_SIPHASH
 
+#define EXT4_MAX_CHECKSUM_SIZE	4
+
 static inline u32 ext4_chksum(struct ext4_sb_info *sbi, u32 crc,
 			      const void *address, unsigned int length)
 {
-	struct {
-		struct shash_desc shash;
-		char ctx[4];
-	} desc;
+	DEFINE_RAW_FLEX(struct shash_desc, desc, __ctx,
+		DIV_ROUND_UP(EXT4_MAX_CHECKSUM_SIZE,
+			     sizeof(*((struct shash_desc *)0)->__ctx)));
 
-	BUG_ON(crypto_shash_descsize(sbi->s_chksum_driver)!=sizeof(desc.ctx));
+	BUG_ON(crypto_shash_descsize(sbi->s_chksum_driver) !=
+		EXT4_MAX_CHECKSUM_SIZE);
 
-	desc.shash.tfm = sbi->s_chksum_driver;
-	*(u32 *)desc.ctx = crc;
+	desc->tfm = sbi->s_chksum_driver;
+	*(u32 *)desc->__ctx = crc;
 
-	BUG_ON(crypto_shash_update(&desc.shash, address, length));
+	BUG_ON(crypto_shash_update(desc, address, length));
 
-	return *(u32 *)desc.ctx;
+	return *(u32 *)desc->__ctx;
 }
 
 #ifdef __KERNEL__
-- 
2.43.0

Re: [PATCH][next] ext4: avoid dozens of -Wflex-array-member-not-at-end warnings

[Theodore Ts'o]

On Tue, Nov 19, 2024 at 04:37:05PM -0600, Gustavo A. R. Silva wrote:
> -Wflex-array-member-not-at-end was introduced in GCC-14, and we
> are getting ready to enable it, globally.
> 
> Use the `DEFINE_RAW_FLEX()` helper for an on-stack definition of
> a flexible structure (`struct shash_desc`) where the size of the
> flexible-array member (`__ctx`) is known at compile-time, and
> refactor the rest of the code, accordingly.
> 
> So, with these changes, fix 38 of the following warnings:
> 
> fs/ext4/ext4.h:2471:35: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
> 
> Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>

Thanks, for this patch!  It appears that this patch has since been
obviated by Eric Bigger's commit f2b4fa19647e (" ext4: switch to using
the crc32c library"), which landed during this merge window, so this
patch should not be needed.

						- Ted

Re: [PATCH][next] ext4: avoid dozens of -Wflex-array-member-not-at-end warnings

[Gustavo A. R. Silva]

> Thanks, for this patch!  It appears that this patch has since been
> obviated by Eric Bigger's commit f2b4fa19647e (" ext4: switch to using
> the crc32c library"), which landed during this merge window, so this
> patch should not be needed.

Oh, nice! Thanks for letting me know. :)

-Gustavo

Re: [PATCH][next] ext4: avoid dozens of -Wflex-array-member-not-at-end warnings

[Theodore Ts'o]

On Fri, Feb 07, 2025 at 09:39:42AM +1030, Gustavo A. R. Silva wrote:
> 
> > Thanks, for this patch!  It appears that this patch has since been
> > obviated by Eric Bigger's commit f2b4fa19647e (" ext4: switch to using
> > the crc32c library"), which landed during this merge window, so this
> > patch should not be needed.
> 
> Oh, nice! Thanks for letting me know. :)

Yeah, apologiues for not getting back to you earlier.  Between travel
and the holidays, I didn't have time to process patches for the merge
window which just closed, so I'm not catching up on the backlog.

In any case, Eric's commit is in 6.14-rc1, so you should see the flex
array warnings from the ext4 checksum inline function.

       	     	  	     	     - Ted