This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCHv2] powerpc: Add optimized P8 strspn
- From: "Paul E. Murphy" <murphyp at linux dot vnet dot ibm dot com>
- To: Segher Boessenkool <segher at kernel dot crashing dot org>
- Cc: "libc-alpha at sourceware dot org" <libc-alpha at sourceware dot org>, Tulio Magno Quites Machado Filho <tuliom at linux dot vnet dot ibm dot com>
- Date: Mon, 4 Apr 2016 09:48:46 -0500
- Subject: Re: [PATCHv2] powerpc: Add optimized P8 strspn
- Authentication-results: sourceware.org; auth=none
- References: <56FC0008 dot 7030302 at linux dot vnet dot ibm dot com> <20160402234040 dot GA17716 at gate dot crashing dot org>
On 04/02/2016 06:40 PM, Segher Boessenkool wrote:
> Hi Paul,
>
> Sorry for not reviewing this earlier. Some minor things; this is
> not a full review.
>
>> +++ b/sysdeps/powerpc/powerpc64/multiarch/strspn.c
>> @@ -0,0 +1,35 @@
>> +/* Multiple versions of strspn. PowerPC64 version.
>
> Two spaces after full stop.
>
>> +++ b/sysdeps/powerpc/powerpc64/power8/strspn.S
>
>> +/* size_t [r3] strspn (const char *string [r3],
>> + const char *needleAccept [r4] */
>
> Missing ")".
ack.
>
>> + lvsr v11, r0, r3
>
> That is 0, not r0.
>
>> + lvx v0, r0, r3 /* Note, unaligned load ignores lower bits. */
>
> Same for lvx.
I'd intentionally used r0 in those cases to explicitly exploit its special
meaning under those instructions. It it customary to explicitly use 0 in
such cases?
Thank you for reviewing!
-Paul