https://sourceware.org/bugzilla/show_bug.cgi?id=32403
Haochen Jiang changed:
What|Removed |Added
Resolution|--- |FIXED
Status|NEW
https://sourceware.org/bugzilla/show_bug.cgi?id=32403
--- Comment #9 from Haochen Jiang ---
Proposed patch:
https://sourceware.org/pipermail/binutils/2024-December/137954.html
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=32403
--- Comment #7 from Haochen Jiang ---
(In reply to H.J. Lu from comment #5)
> Created attachment 15811 [details]
> There are more.
Thanks!
I would prefer the name as something like ME not Xe to avoid confusion for
lower case letters in mnemo
https://sourceware.org/bugzilla/show_bug.cgi?id=32403
--- Comment #4 from Haochen Jiang ---
(In reply to Haochen Jiang from comment #3)
> I will have a look into that.
>
> From my memory in assembler implementation, the {evex} prefix relies on and
> only on the register usage. It won't check mem
https://sourceware.org/bugzilla/show_bug.cgi?id=32403
--- Comment #3 from Haochen Jiang ---
I will have a look into that.
>From my memory in assembler implementation, the {evex} prefix relies on and
only on the register usage. It won't check memory and will not consider whether
an instruction wi
https://sourceware.org/bugzilla/show_bug.cgi?id=31043
Haochen Jiang changed:
What|Removed |Added
CC||haochen.jiang at intel dot com
https://sourceware.org/bugzilla/show_bug.cgi?id=31043
--- Comment #1 from Haochen Jiang ---
Maybe we have a chance to emit an error for invalid_register_operand in
check_VecOperands under this scenario. But I am not sure if we really need
that.
BTW, this error type seems not used anymore. We cou
https://sourceware.org/bugzilla/show_bug.cgi?id=30317
--- Comment #3 from Haochen Jiang ---
(In reply to Jan Beulich from comment #2)
> SwapSources can only sensibly be present / in use for real insn templates.
Yes since when we are in s_insn(), we actually did not know whether it is swap
source
https://sourceware.org/bugzilla/show_bug.cgi?id=30317
--- Comment #1 from Haochen Jiang ---
I see why testsuite did not report this is an error since the insn code did
generate the corresponding output expected. We just only need to solve the
reversed operand problem.
--
You are receiving this
https://sourceware.org/bugzilla/show_bug.cgi?id=30317
Haochen Jiang changed:
What|Removed |Added
CC||hjl.tools at gmail dot com,
Component: gas
Assignee: unassigned at sourceware dot org
Reporter: haochen.jiang at intel dot com
Target Milestone: ---
For the current AMX and AMX-FP16 bad testcases, we got testcases like this:
#tdpfp16ps %tmm5,%tmm4,%tmm3 set VEX.W = 1 (illegal value).
.insn VEX
11 matches
Mail list logo
