This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: Bug: gather/scatter accepts same registers in destination/src and index
- From: "Jan Beulich" <JBeulich at suse dot com>
- To: "Michael Zolotukhin" <michael dot v dot zolotukhin at gmail dot com>
- Cc: "H.J. Lu" <hjl dot tools at gmail dot com>, "binutils at sourceware dot org" <binutils at sourceware dot org>
- Date: Tue, 21 Jan 2014 15:32:42 +0000
- Subject: Re: Bug: gather/scatter accepts same registers in destination/src and index
- Authentication-results: sourceware.org; auth=none
- References: <CANtU078XuYEQ8AnKR2BQoRSSByKLAf_TrzdpjEO=4h09U524gQ at mail dot gmail dot com> <52DE438B0200007800115421 at nat28 dot tlf dot novell dot com> <20140121111627 dot GC13904 at msticlxl57 dot ims dot intel dot com> <CAMe9rOpLKpmSR-vYpOE5RS3BDn2P7PewSVh85-zTHy4ZUxXvuA at mail dot gmail dot com> <20140121135628 dot GF13904 at msticlxl57 dot ims dot intel dot com> <52DE93F00200007800115670 at nat28 dot tlf dot novell dot com> <CANtU0782E-KwJbEQf=Otvq0h+0vhLXoi2yYip0gkZdonRBVvOQ at mail dot gmail dot com> <52DE96420200007800115697 at nat28 dot tlf dot novell dot com> <CANtU07_huE8K_xvb9LNg11F9zZ5RzGO9gT_r=b9j_DVG3y1PzA at mail dot gmail dot com>
>>> On 21.01.14 at 16:09, Michael Zolotukhin <michael.v.zolotukhin@gmail.com> wrote:
>> I was talking of the test case here (i.e. validate that scatters without
>> index get warned/errored upon).
> How is it different from any other instruction with incorrect operands?
> Could
> you give an example of a test you mean?
While
vmovaps [rax], zmm0
is allowed
vscatterdps [rax]{k1}, zmm0
isn't, i.e. this is not a typical "invalid operand". Same for gather
of course. Also, I'm not sure if this
vscatterdps [rax+zmm1], zmm0
(also invalid iirc) is being checked for already.
And your argument is bogus for another reason too: If so,
checking register collisions for gather to get properly detected
would then also be "superfluous".
Jan