This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [SH] Correct fabs and fneg insns in simulator
- From: Joel Brobecker <brobecker at adacore dot com>
- To: Oleg Endo <oleg dot endo at t-online dot de>
- Cc: gdb-patches at sourceware dot org, Kaz Kojima <kkojima at rr dot iij4u dot or dot jp>
- Date: Fri, 28 Nov 2014 19:51:22 +0400
- Subject: Re: [SH] Correct fabs and fneg insns in simulator
- Authentication-results: sourceware.org; auth=none
- References: <1413243560 dot 31394 dot 48 dot camel at yam-132-YW-E178-FTW>
Hello Oleg,,
> It seems that the implementation of the SH fabs and fneg insns in the
> simulator is not correct. They use the FP_UNARY macro which checks the
> FPSCR.PR setting and raises an exception if PR = 1 (double precision)
> and the register number is not even (i.e. a valid DF reg number).
> For normal unary FP insns this is fine. However, fneg and fabs perform
> the same (integer) operations regardless of the FPSCR.PR setting.
>
> This issue initially popped up here
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63260
>
> I've checked some of the failing tests mentioned in GCC PR 63260 above
> with the patch applied and the failures go away.
>
> Could somebody please apply it if it's OK?
>
> I'm not subscribed to gdb-patches, please CC me when replying.
>
> Cheers,
> Oleg
>
> sim/ChangeLog:
>
> 2014-10-14 Oleg Endo <olegendo@gcc.gnu.org>
>
> * sh/gencode.c (fabs, fneg): Implement as integer operation
> instead of using the FP_UNARY macro.
Thank you for the patch. We don't have an SH maintainer for GDB,
and in the absence of any feedback from other interested people,
we will just trust you on this. So I pushed this patch under the
"tiny patch" rule.
If you ever think of contributing other patches to GDB, we'll probably
have to start thinking about getting your copyright assignment to
cover GDB as well.
--
Joel