This is the mail archive of the libc-hacker@sources.redhat.com mailing list for the glibc project.

Note that libc-hacker is a closed list. You may look at the archives of this list, but subscription and posting are not open.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [PATCH] PPC enable Alitvec hardware, part 1

From: Steve Munroe <sjmunroe at us dot ibm dot com>
To: libc-hacker at sources dot redhat dot com
Cc: Benjamin Herrenschmidt <benh at kernel dot crashing dot org>, Paul Mackerras <paulus at samba dot org>
Date: Sun, 11 Jan 2004 10:41:52 -0600
Subject: Re: [PATCH] PPC enable Alitvec hardware, part 1

Roland McGrath writes:

> > The attached patch provides base enabling for PowerPC Altivec (AKA 
VMX) 
> > and for versioning effected functions. 2.6.0 kernels provide a 
> > PPC_FEATURE_HAS_ALTIVEC status in the AT_HWCAP. We need to capture 
this 
> > status as a static flag (__has_altivec) so functions enabled for 
altivec 
> > can run safely on PowerPC hardware without the Altivec feature.
>
> Why did you introduce this variable?  Using GL(dl_hwcap) should be fine.

GL(dl_hwcap) is not very asm friendly. ldsodefs.h is full of things that 
gas does not like. Plus in the shared case _rtld_global.dl_hwcap is a 
struct/field and there is no (automatic/correct) way to obtain the offset 
of the dl_hwcap field for the asm code. 

Steven J. Munroe
Power Linux Toolchain Architect
IBM Corporation, Linux Technology Center

Follow-Ups:
- Re: [PATCH] PPC enable Alitvec hardware, part 1
  - From: Ulrich Drepper

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]