From: Jeff Law <jeffreyalaw@gmail.com>
To: Jin Ma <jinma@linux.alibaba.com>, gcc-patches <gcc-patches@gcc.gnu.org>
Cc: palmer <palmer@dabbelt.com>,
"richard.sandiford" <richard.sandiford@arm.com>,
"kito.cheng" <kito.cheng@gmail.com>,
"philipp.tomsich" <philipp.tomsich@vrull.eu>,
"christoph.muellner" <christoph.muellner@vrull.eu>,
"rdapp.gcc" <rdapp.gcc@gmail.com>,
"juzhe.zhong" <juzhe.zhong@rivai.ai>,
"jinma.contrib" <jinma.contrib@gmail.com>
Subject: Re: [RFC 1/2] RISC-V: Add support for _Bfloat16.
Date: Mon, 9 Oct 2023 13:16:04 -0600 [thread overview]
Message-ID: <06e35f34-2301-4a60-8dae-797925e88c0c@gmail.com> (raw)
In-Reply-To: <be4da547-5ef5-4c54-8dc8-26a7ff9b48c5.jinma@linux.alibaba.com>
On 10/9/23 00:18, Jin Ma wrote:
>>> +;; The conversion of DF to BF needs to be done with SF if there is a
>>> +;; chance to generate at least one instruction, otherwise just using
>>> +;; libfunc __truncdfbf2.
>>> +(define_expand "truncdfbf2"
>>> + [(set (match_operand:BF 0 "register_operand" "=f")
>>> + (float_truncate:BF
>>> + (match_operand:DF 1 "register_operand" " f")))]
>>> + "TARGET_DOUBLE_FLOAT || TARGET_ZDINX"
>>> + {
>>> + convert_move (operands[0],
>>> + convert_modes (SFmode, DFmode, operands[1], 0), 0);
>>> + DONE;
>>> + })
>> So for conversions to/from BFmode, doesn't generic code take care of
>> this for us? Search for convert_mode_scalar in expr.cc. That code will
>> utilize SFmode as an intermediate step just like your expander. Is
>> there some reason that generic code is insufficient?
>>
>> Similarly for the the other conversions.
>
> As far as I can see, the function 'convert_mode_scalar' doesn't seem to be perfect for
> dealing with the conversions to/from BFmode. It can only handle BF to HF, SF, DF and
> SF to BF well, but the rest of the conversion without any processing, directly using
> the libcall.
>
> Maybe I should choose to enhance its functionality? This seems to be a
> good choice, I'm not sure.My recollection was that BF could be converted to/from SF trivially and
if we wanted BF->DF we'd first convert to SF, then to DF.
Direct BF<->DF conversions aren't actually important from a performance
standpoint. So it's OK if they have an extra step IMHO.
jeff
next prev parent reply other threads:[~2023-10-09 19:16 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-19 8:44 Jin Ma
2023-09-19 8:46 ` [RFC 2/2] RISC-V: Add 'Zfbfmin' extension Jin Ma
2023-09-29 17:52 ` Jeff Law
2023-09-29 17:46 ` [RFC 1/2] RISC-V: Add support for _Bfloat16 Jeff Law
2023-10-09 6:18 ` Jin Ma
2023-10-09 19:16 ` Jeff Law [this message]
2023-10-25 10:15 ` Jin Ma
2023-11-10 21:21 ` Jeff Law
2023-12-20 11:17 ` [PATCH] Support libcall __float{,un}sibf by SF when it is not supported for _bf16 Jin Ma
2024-01-09 9:32 ` Jin Ma
2024-01-09 9:43 ` Jin Ma
2024-05-26 14:53 ` [PATCH] " Jeff Law
2024-06-05 6:18 ` Jin Ma
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=06e35f34-2301-4a60-8dae-797925e88c0c@gmail.com \
--to=jeffreyalaw@gmail.com \
--cc=christoph.muellner@vrull.eu \
--cc=gcc-patches@gcc.gnu.org \
--cc=jinma.contrib@gmail.com \
--cc=jinma@linux.alibaba.com \
--cc=juzhe.zhong@rivai.ai \
--cc=kito.cheng@gmail.com \
--cc=palmer@dabbelt.com \
--cc=philipp.tomsich@vrull.eu \
--cc=rdapp.gcc@gmail.com \
--cc=richard.sandiford@arm.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).