Re: [PATCH 5.4 243/321] x86/asm/crypto: Annotate local functions

[Eric Biggers <ebiggers@kernel.org>]

On Thu, Dec 12, 2024 at 07:05:24PM +0100, Greg Kroah-Hartman wrote:
> On Thu, Dec 12, 2024 at 06:00:23PM +0000, Eric Biggers wrote:
> > On Thu, Dec 12, 2024 at 04:02:41PM +0100, Greg Kroah-Hartman wrote:
> > > 5.4-stable review patch.  If anyone has any objections, please let me know.
> > > 
> > > ------------------
> > > 
> > > From: Jiri Slaby <jslaby@suse.cz>
> > > 
> > > [ Upstream commit 74d8b90a889022e306b543ff2147a6941c99b354 ]
> > > 
> > > Use the newly added SYM_FUNC_START_LOCAL to annotate beginnings of all
> > > functions which do not have ".globl" annotation, but their endings are
> > > annotated by ENDPROC. This is needed to balance ENDPROC for tools that
> > > generate debuginfo.
> > > 
> > > These function names are not prepended with ".L" as they might appear in
> > > call traces and they wouldn't be visible after such change.
> > > 
> > > To be symmetric, the functions' ENDPROCs are converted to the new
> > > SYM_FUNC_END.
> > > 
> > > Signed-off-by: Jiri Slaby <jslaby@suse.cz>
> > > Signed-off-by: Borislav Petkov <bp@suse.de>
> > > Cc: "David S. Miller" <davem@davemloft.net>
> > > Cc: Herbert Xu <herbert@gondor.apana.org.au>
> > > Cc: "H. Peter Anvin" <hpa@zytor.com>
> > > Cc: Ingo Molnar <mingo@redhat.com>
> > > Cc: linux-arch@vger.kernel.org
> > > Cc: linux-crypto@vger.kernel.org
> > > Cc: Thomas Gleixner <tglx@linutronix.de>
> > > Cc: x86-ml <x86@kernel.org>
> > > Link: https://lkml.kernel.org/r/20191011115108.12392-7-jslaby@suse.cz
> > > Stable-dep-of: 3b2f2d22fb42 ("crypto: x86/aegis128 - access 32-bit arguments as 32-bit")
> > > Signed-off-by: Sasha Levin <sashal@kernel.org>
> > > ---
> > >  arch/x86/crypto/aegis128-aesni-asm.S         |  8 ++--
> > >  arch/x86/crypto/aesni-intel_asm.S            | 49 ++++++++------------
> > >  arch/x86/crypto/camellia-aesni-avx-asm_64.S  | 20 ++++----
> > >  arch/x86/crypto/camellia-aesni-avx2-asm_64.S | 20 ++++----
> > >  arch/x86/crypto/cast5-avx-x86_64-asm_64.S    |  8 ++--
> > >  arch/x86/crypto/cast6-avx-x86_64-asm_64.S    |  8 ++--
> > >  arch/x86/crypto/chacha-ssse3-x86_64.S        |  4 +-
> > >  arch/x86/crypto/ghash-clmulni-intel_asm.S    |  4 +-
> > >  arch/x86/crypto/serpent-avx-x86_64-asm_64.S  |  8 ++--
> > >  arch/x86/crypto/serpent-avx2-asm_64.S        |  8 ++--
> > >  arch/x86/crypto/twofish-avx-x86_64-asm_64.S  |  8 ++--
> > >  11 files changed, 68 insertions(+), 77 deletions(-)
> > 
> > Unless the author of this patch acks this I'd rather you skipped this.  It's not
> > worth the risk of regressions in the crypto code.
> 
> It's a dependancy of commit 3b2f2d22fb42 ("crypto: x86/aegis128 - access
> 32-bit arguments as 32-bit"), so should we drop that one also?
> 

Well it is not a dependency if the conflict is properly resolved, but I would
just drop it too.  In theory it fixes a bug, but we haven't seen gcc or clang
generating code that makes it matter.  Also I've noticed that some other asm
files have the same issue...

- Eric