linux-kernel - Re: [PATCH] X86-32: Let gcc decide whether to inline memcpy was Re: New x86 warning

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 23 Apr 2009 14:22:02 -0700
From:	Joe Damato <ice799@...il.com>
To:	"H. Peter Anvin" <hpa@...or.com>
Cc:	Andi Kleen <andi@...stfloor.org>, Ingo Molnar <mingo@...e.hu>,
	Jeff Garzik <jeff@...zik.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	LKML <linux-kernel@...r.kernel.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Ingo Molnar <mingo@...hat.com>, x86@...nel.org
Subject: Re: [PATCH] X86-32: Let gcc decide whether to inline memcpy was Re: 
	New x86 warning

On Wed, Apr 22, 2009 at 6:48 PM, H. Peter Anvin <hpa@...or.com> wrote:
> Joe Damato wrote:
>>
>> I think this patch is great. Perhaps this would be a good time to also
>> clean out memset for x86_32? (If needed, I can start a new email
>> thread with this patch) I've built and booted this on my x86_32
>> hardware.
>>
>
> Under the same conditions, i.e. making sure that with the oldest gcc we
> still care about (3.2) we don't end up uninlining any new cases we
> shouldn't.

Looks like this thread is dead/dying, but figured I should reply with
my test findings. The number of out-of-line calls (as determined by:
make mrproper && make defconfig && make && objdump -d vmlinux | grep
"call.*\<memset" | wc -l))

gcc 4.2.4 - withOUT memset patch: 20
gcc 4.2.4 - with memset patch: 365

gcc 3.4 - withOUT memset patch: 17
gcc 3.4 - with memset patch: 349

I'm guessing this is probably not acceptable, so I won't bother
installing/trying gcc-3.2 unless anyone thinks that a 300+ increase in
out-of-line calls is OK.

Joe
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/