From: Richard Biener <richard.guenther@gmail.com>
To: Segher Boessenkool <segher@kernel.crashing.org>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH 0/5] RFC, WIP: RTL cost improvements
Date: Tue, 01 Aug 2017 09:07:00 -0000 [thread overview]
Message-ID: <CAFiYyc0gD0KVbJK7e-+WS_q7kjCmXfGJB+q_BGe8CsJ0rnubuA@mail.gmail.com> (raw)
In-Reply-To: <cover.1501541006.git.segher@kernel.crashing.org>
On Tue, Aug 1, 2017 at 1:24 AM, Segher Boessenkool
<segher@kernel.crashing.org> wrote:
> This series creates pattern_cost and insn_cost functions that together
> replace the existing insn_rtx_cost function.
>
> pattern_cost is like the old insn_rtx_cost function; insn_cost takes
> an actual rtx_insn * as input, not just a pattern.
>
> Also a targetm.insn_cost is added, which targets can use to implement
> a more exact cost more easily.
>
> The combine patch is pretty gross (but functional), it needs some
> refactoring (to not call recog so often). The rs6000 patch is very
> much a work in progress.
>
> How does this look? Is this the right direction?
I think it is the right direction. I'll leave the details to somebody more
familiar with RTL.
Richard.
>
> Segher
>
>
> Segher Boessenkool (5):
> Rename existing insn_cost to insn_sched_cost
> Replace insn_rtx_cost with insn_cost and pattern_cost
> combine: Use insn_cost instead of pattern_cost everywhere
> Add targetm.insn_cost hook
> rs6000: Implement insn_cost hook
>
> gcc/cfgrtl.c | 7 +++----
> gcc/combine.c | 40 +++++++++++++++++++++++++++++-----------
> gcc/config/rs6000/rs6000.c | 14 ++++++++++++++
> gcc/doc/tm.texi | 12 ++++++++++++
> gcc/doc/tm.texi.in | 2 ++
> gcc/dse.c | 2 +-
> gcc/haifa-sched.c | 14 +++++++-------
> gcc/ifcvt.c | 12 ++++++------
> gcc/rtl.h | 3 ++-
> gcc/rtlanal.c | 16 ++++++++++++++--
> gcc/sched-int.h | 2 +-
> gcc/sched-rgn.c | 4 ++--
> gcc/sel-sched-ir.c | 2 +-
> gcc/target.def | 14 ++++++++++++++
> 14 files changed, 108 insertions(+), 36 deletions(-)
>
> --
> 1.9.3
>
next prev parent reply other threads:[~2017-08-01 9:07 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-01 0:15 Segher Boessenkool
2017-08-01 0:05 ` [PATCH 4/5] Add targetm.insn_cost hook Segher Boessenkool
2017-08-02 16:39 ` Jeff Law
2017-08-01 0:05 ` [PATCH 3/5] combine: Use insn_cost instead of pattern_cost everywhere Segher Boessenkool
2017-08-02 16:40 ` Jeff Law
2017-08-02 21:23 ` Segher Boessenkool
2017-08-01 0:05 ` [PATCH 2/5] Replace insn_rtx_cost with insn_cost and pattern_cost Segher Boessenkool
2017-08-02 16:38 ` Jeff Law
2017-08-01 0:15 ` [PATCH 5/5] rs6000: Implement insn_cost hook Segher Boessenkool
2017-08-01 0:15 ` [PATCH 1/5] Rename existing insn_cost to insn_sched_cost Segher Boessenkool
2017-08-02 16:36 ` Jeff Law
2017-08-01 9:07 ` Richard Biener [this message]
2017-08-05 10:07 ` [PATCH 0/5] RFC, WIP: RTL cost improvements Richard Sandiford
2017-08-05 17:15 ` Segher Boessenkool
2017-08-08 16:41 ` Jeff Law
2017-08-08 16:55 ` Richard Sandiford
2017-08-09 15:56 ` Jeff Law
2017-08-09 16:19 ` Richard Sandiford
2017-08-09 16:40 ` Segher Boessenkool
2017-08-09 16:54 ` Richard Sandiford
2017-08-09 17:12 ` Segher Boessenkool
2017-08-09 18:18 ` Segher Boessenkool
2017-08-10 7:18 ` Richard Sandiford
2017-08-09 16:57 ` Segher Boessenkool
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=CAFiYyc0gD0KVbJK7e-+WS_q7kjCmXfGJB+q_BGe8CsJ0rnubuA@mail.gmail.com \
--to=richard.guenther@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=segher@kernel.crashing.org \
/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).