From: Jan Beulich <jbeulich@suse.com>
To: "Cui, Lili" <lili.cui@intel.com>
Cc: "Lu, Hongjiu" <hongjiu.lu@intel.com>,
"binutils@sourceware.org" <binutils@sourceware.org>
Subject: Re: [PATCH v3 7/9] Support APX Push2/Pop2
Date: Fri, 15 Dec 2023 09:44:09 +0100 [thread overview]
Message-ID: <8f935a02-5270-4467-9741-3edbf2a41ba1@suse.com> (raw)
In-Reply-To: <SJ0PR11MB56009090E12FAC65DA348EFA9E93A@SJ0PR11MB5600.namprd11.prod.outlook.com>
On 15.12.2023 09:38, Cui, Lili wrote:
>> On 24.11.2023 08:02, Cui, Lili wrote:
>>> @@ -5422,6 +5427,9 @@ md_assemble (char *line)
>>> case unsupported_rc_sae:
>>> err_msg = _("unsupported static rounding/sae");
>>> break;
>>> + case unsupported_rsp_register:
>>> + err_msg = _("cannot be used with %rsp register");
>>> + break;
>>
>> While this wording looks okay as visible here, please consider it in the context
>> it is used in: "cannot be used with %rsp register for `push2'"
>> is, I'm sorry to say that, clumsy at best. If you want to stick to setting err_msg,
>> how about "%rsp register cannot be used"? Personally I'd prefer a resulting
>> output of "%rsp register cannot be used with `push2'", but I wouldn't insist on
>> you going that route if you don't like that.
>
> "%rsp register cannot be used" ,this is much better, thanks.
It occurs to me only now (sorry) that %rsp is inappropriate to use
when assembling Intel syntax insns. In that case the % may not be there
(and as a result the entire register name then wants putting in single
quotes, as we do elsewhere).
>>> +check_APX_operands (const insn_template *t) {
>>> + /* Push2* and Pop2* cannot use RSP and Pop2* cannot pop two same
>> registers.
>>> + */
>>> + if (t->mnem_off == MN_push2 || t->mnem_off == MN_push2p
>>> + || t->mnem_off == MN_pop2 || t->mnem_off == MN_pop2p)
>>
>> Considering (perhaps just theoretical) further additions here, did you consider
>> using switch()? Even without further additions this would imo be more legible
>> (due to there being slightly less redundancy).
>>
>
> Done.
>
> /* Push2* and Pop2* cannot use RSP and Pop2* cannot pop two same registers.
> */
> switch (t->mnem_off)
> {
> case MN_pop2:
> case MN_pop2p:
> if (register_number (i.op[0].regs) == register_number (i.op[1].regs))
> {
> i.error = invalid_dest_register_set;
> return 1;
> }
Fall-through comment here please, ...
> case MN_push2:
> case MN_push2p:
> if (register_number (i.op[0].regs) == 4
> || register_number (i.op[1].regs) == 4)
> {
> i.error = unsupported_rsp_register;
> return 1;
> }
... and break here please.
Jan
next prev parent reply other threads:[~2023-12-15 8:44 UTC|newest]
Thread overview: 69+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-24 7:02 [PATCH 1/9] Make const_1_mode print $1 in AT&T syntax Cui, Lili
2023-11-24 7:02 ` [PATCH v3 2/9] Support APX GPR32 with rex2 prefix Cui, Lili
2023-12-04 16:30 ` Jan Beulich
2023-12-05 13:31 ` Cui, Lili
2023-12-06 7:52 ` Jan Beulich
2023-12-06 12:43 ` Cui, Lili
2023-12-07 9:01 ` Jan Beulich
2023-12-08 3:10 ` Cui, Lili
2023-11-24 7:02 ` [PATCH v3 3/9] Created an empty EVEX_MAP4_ sub-table for EVEX instructions Cui, Lili
2023-11-24 7:02 ` [PATCH v3 4/9] Support APX GPR32 with extend evex prefix Cui, Lili
2023-12-07 12:38 ` Jan Beulich
2023-12-08 15:21 ` Cui, Lili
2023-12-11 8:34 ` Jan Beulich
2023-12-12 10:44 ` Cui, Lili
2023-12-12 11:16 ` Jan Beulich
2023-12-12 12:32 ` Cui, Lili
2023-12-12 12:39 ` Jan Beulich
2023-12-12 13:15 ` Cui, Lili
2023-12-12 14:13 ` Jan Beulich
2023-12-13 7:36 ` Cui, Lili
2023-12-13 7:48 ` Jan Beulich
2023-12-12 12:58 ` Cui, Lili
2023-12-12 14:04 ` Jan Beulich
2023-12-13 8:35 ` Cui, Lili
2023-12-13 9:13 ` Jan Beulich
2023-12-07 13:34 ` Jan Beulich
2023-12-11 6:16 ` Cui, Lili
2023-12-11 8:43 ` Jan Beulich
2023-12-11 11:50 ` Jan Beulich
2023-11-24 7:02 ` [PATCH v3 5/9] Add tests for " Cui, Lili
2023-12-07 14:05 ` Jan Beulich
2023-12-11 6:16 ` Cui, Lili
2023-12-11 8:55 ` Jan Beulich
2023-11-24 7:02 ` [PATCH v3 6/9] Support APX NDD Cui, Lili
2023-12-08 14:12 ` Jan Beulich
2023-12-11 13:36 ` Cui, Lili
2023-12-11 16:50 ` Jan Beulich
2023-12-13 10:42 ` Cui, Lili
2024-03-22 10:02 ` Jan Beulich
2024-03-22 10:31 ` Jan Beulich
2024-03-26 2:04 ` Cui, Lili
2024-03-26 7:06 ` Jan Beulich
2024-03-26 7:18 ` Cui, Lili
2024-03-22 10:59 ` Jan Beulich
2024-03-26 8:22 ` Cui, Lili
2024-03-26 9:30 ` Jan Beulich
2024-03-27 2:41 ` Cui, Lili
2023-12-08 14:27 ` Jan Beulich
2023-12-12 5:53 ` Cui, Lili
2023-12-12 8:28 ` Jan Beulich
2023-11-24 7:02 ` [PATCH v3 7/9] Support APX Push2/Pop2 Cui, Lili
2023-12-11 11:17 ` Jan Beulich
2023-12-15 8:38 ` Cui, Lili
2023-12-15 8:44 ` Jan Beulich [this message]
2023-11-24 7:02 ` [PATCH v3 8/9] Support APX NDD optimized encoding Cui, Lili
2023-12-11 12:27 ` Jan Beulich
2023-12-12 3:18 ` Hu, Lin1
2023-12-12 8:41 ` Jan Beulich
2023-12-13 5:31 ` Hu, Lin1
2023-12-12 8:45 ` Jan Beulich
2023-12-13 6:06 ` Hu, Lin1
2023-12-13 8:19 ` Jan Beulich
2023-12-13 8:34 ` Hu, Lin1
2023-11-24 7:02 ` [PATCH v3 9/9] Support APX JMPABS for disassembler Cui, Lili
2023-11-24 7:09 ` [PATCH 1/9] Make const_1_mode print $1 in AT&T syntax Jan Beulich
2023-11-24 11:22 ` Cui, Lili
2023-11-24 12:14 ` Jan Beulich
2023-12-12 2:57 ` Lu, Hongjiu
2023-12-12 8:16 ` Cui, Lili
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=8f935a02-5270-4467-9741-3edbf2a41ba1@suse.com \
--to=jbeulich@suse.com \
--cc=binutils@sourceware.org \
--cc=hongjiu.lu@intel.com \
--cc=lili.cui@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).