| Message ID | ZD+0DJq1NHmMSSja@gondor.apana.org.au (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Herbert Xu |
| Headers | show |
| Series | crypto: arm/sha1-neon - Fix clang function cast warnings | expand |
On Wed, Apr 19, 2023 at 2:27 AM Herbert Xu <herbert@gondor.apana.org.au> wrote: > > Instead of casting the function which upsets clang for some reason, > change the assembly function siganture instead. s/siganture/signature/ Thanks for the patch! Perhaps worth noting in the commit message that this fixes an instance of -Wcast-function-type-strict triggered for W=1 builds, which helps us catch potential kCFI violations at compile time rather than runtime. After this patch, there's only 4 more users of the sha1_block_fn typedef, consider if that typedef is a candidate for removal now? Also, you should update the comment in arch/arm/crypto/sha1-armv7-neon.S above the assembler implementation on L288-289 to reflect this revised fn signature, in this patch. > > Reported-by: kernel test robot <lkp@intel.com> > Link: https://lore.kernel.org/oe-kbuild-all/202304081828.zjGcFUyE-lkp@intel.com/ > Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> > > diff --git a/arch/arm/crypto/sha1_neon_glue.c b/arch/arm/crypto/sha1_neon_glue.c > index cfe36ae0f3f5..9c70b87e69f7 100644 > --- a/arch/arm/crypto/sha1_neon_glue.c > +++ b/arch/arm/crypto/sha1_neon_glue.c > @@ -26,8 +26,8 @@ > > #include "sha1.h" > > -asmlinkage void sha1_transform_neon(void *state_h, const char *data, > - unsigned int rounds); > +asmlinkage void sha1_transform_neon(struct sha1_state *state_h, > + const u8 *data, int rounds); > > static int sha1_neon_update(struct shash_desc *desc, const u8 *data, > unsigned int len) > @@ -39,8 +39,7 @@ static int sha1_neon_update(struct shash_desc *desc, const u8 *data, > return sha1_update_arm(desc, data, len); > > kernel_neon_begin(); > - sha1_base_do_update(desc, data, len, > - (sha1_block_fn *)sha1_transform_neon); > + sha1_base_do_update(desc, data, len, sha1_transform_neon); > kernel_neon_end(); > > return 0; > @@ -54,9 +53,8 @@ static int sha1_neon_finup(struct shash_desc *desc, const u8 *data, > > kernel_neon_begin(); > if (len) > - sha1_base_do_update(desc, data, len, > - (sha1_block_fn *)sha1_transform_neon); > - sha1_base_do_finalize(desc, (sha1_block_fn *)sha1_transform_neon); > + sha1_base_do_update(desc, data, len, sha1_transform_neon); > + sha1_base_do_finalize(desc, sha1_transform_neon); > kernel_neon_end(); > > return sha1_base_finish(desc, out); > -- > Email: Herbert Xu <herbert@gondor.apana.org.au> > Home Page: http://gondor.apana.org.au/~herbert/ > PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt >
diff --git a/arch/arm/crypto/sha1_neon_glue.c b/arch/arm/crypto/sha1_neon_glue.c index cfe36ae0f3f5..9c70b87e69f7 100644 --- a/arch/arm/crypto/sha1_neon_glue.c +++ b/arch/arm/crypto/sha1_neon_glue.c @@ -26,8 +26,8 @@ #include "sha1.h" -asmlinkage void sha1_transform_neon(void *state_h, const char *data, - unsigned int rounds); +asmlinkage void sha1_transform_neon(struct sha1_state *state_h, + const u8 *data, int rounds); static int sha1_neon_update(struct shash_desc *desc, const u8 *data, unsigned int len) @@ -39,8 +39,7 @@ static int sha1_neon_update(struct shash_desc *desc, const u8 *data, return sha1_update_arm(desc, data, len); kernel_neon_begin(); - sha1_base_do_update(desc, data, len, - (sha1_block_fn *)sha1_transform_neon); + sha1_base_do_update(desc, data, len, sha1_transform_neon); kernel_neon_end(); return 0; @@ -54,9 +53,8 @@ static int sha1_neon_finup(struct shash_desc *desc, const u8 *data, kernel_neon_begin(); if (len) - sha1_base_do_update(desc, data, len, - (sha1_block_fn *)sha1_transform_neon); - sha1_base_do_finalize(desc, (sha1_block_fn *)sha1_transform_neon); + sha1_base_do_update(desc, data, len, sha1_transform_neon); + sha1_base_do_finalize(desc, sha1_transform_neon); kernel_neon_end(); return sha1_base_finish(desc, out);
Instead of casting the function which upsets clang for some reason, change the assembly function siganture instead. Reported-by: kernel test robot <lkp@intel.com> Link: https://lore.kernel.org/oe-kbuild-all/202304081828.zjGcFUyE-lkp@intel.com/ Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>