From: Jeffrey A Law <law@hurl.cygnus.com>
To: "Jerry Quinn" <jquinn@nortelnetworks.com>
Cc: egcs@egcs.cygnus.com
Subject: Re: Fwd: Questions on PA machine description?
Date: Wed, 31 Mar 1999 23:46:00 -0000 [thread overview]
Message-ID: <18375.921819620@hurl.cygnus.com> (raw)
Message-ID: <19990331234600.h4yP3Ov3vHEjSBoFF4Sd449pqW1uCRkjKbBSZyFIFaE@z> (raw)
In-Reply-To: <36F17F8C.9FD937FA@americasm01.nt.com>
In message < 36F17F8C.9FD937FA@americasm01.nt.com >you write:
> The other message mentioned that eliminating autoincrement/autodecrement
> instructions is a good thing. Do these instructions have the same
> problem as fmpyadd, i.e. grabbing multiple reorder slots and function
> units?
Yes, they have the same problem as fmpyadd/fmpysub. They also have the
disadvantage that the autoinc addressing mode adds additional data depedencies
which can inhibit the amount of ILP found by the compiler and by the hardware.
> > One thought would be to make the ready delay for alu instructions 0, then
> > tweak haifa to add dependent instrutions to the ready queue immediately
> > after it issues an insn with a ready delay of zero cycles.
>
> What about making pa_adjust_cost set the cost of a data dependency to
> 0? The alpha port does this on the ev5.
Nope, it won't do what we want. Look at the loop which issues insns from
the ready list in haifa-sched.c.
It has a structure like:
while (not all insns scheduled)
add insns with no outstanding dependencies to the ready queue
sort the ready queue
while (ready list is not empty && target can issue more insns)
issue an insn off the ready queue, remove dependencies on the issued insn
So, given insn1 which feeds insn2 we will never issue insn1 & insn2 in the
same cycle.
> Why is this a good thing? Won't an instruction that depends on another
> one have to retire later than the other one?
No. They can retire in the same cycle. This is discussed in one of the
PA8000 optimization papers from HP. The key is to remember that PA8000
machine is an out of execution machine.
jeff
next prev parent reply other threads:[~1999-03-31 23:46 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
1999-03-08 6:49 Jerry Quinn
[not found] ` < 36E3E35A.4BB420BE@americasm01.nt.com >
1999-03-08 20:48 ` Jeffrey A Law
1999-03-09 11:35 ` Jerry Quinn
[not found] ` < 36E577DF.F825BAFB@americasm01.nt.com >
1999-03-09 12:18 ` Jeffrey A Law
1999-03-16 10:49 ` Jerry Quinn
[not found] ` < 36EEA7B6.D3C894DE@americasm01.nt.com >
1999-03-17 20:12 ` Jeffrey A Law
1999-03-18 14:35 ` Jerry Quinn
[not found] ` < 36F17F8C.9FD937FA@americasm01.nt.com >
1999-03-18 21:00 ` Jeffrey A Law [this message]
[not found] ` < 18375.921819620@hurl.cygnus.com >
1999-03-19 17:32 ` Richard Henderson
[not found] ` < 19990319173226.C14722@cygnus.com >
1999-03-20 2:08 ` Jeffrey A Law
[not found] ` < 1843.921916006@upchuck >
1999-03-20 10:43 ` Richard Henderson
1999-03-23 14:13 ` Jerry Quinn
[not found] ` < 36F811ED.32912006@americasm01.nt.com >
1999-03-24 1:30 ` Jeffrey A Law
1999-03-24 14:44 ` Jerry Quinn
[not found] ` < 36F96A7D.F0DBDF47@americasm01.nt.com >
1999-03-25 1:01 ` Jeffrey A Law
[not found] ` < 4324.922351870@upchuck >
1999-03-25 13:56 ` Richard Henderson
1999-03-25 15:10 ` Richard Henderson
1999-03-26 10:50 ` Jerry Quinn
1999-03-26 11:04 ` Richard Henderson
1999-03-31 23:46 ` Richard Henderson
1999-03-31 23:46 ` Jerry Quinn
1999-03-26 14:07 ` Jerry Quinn
1999-03-27 16:04 ` Jeffrey A Law
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
1999-03-31 23:46 ` Richard Henderson
1999-04-02 11:53 ` Jeffrey A Law
1999-04-05 15:50 ` Jerry Quinn
1999-04-30 23:15 ` Jerry Quinn
1999-04-30 23:15 ` Jeffrey A Law
1999-03-31 23:46 ` Richard Henderson
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
1999-03-31 23:46 ` Richard Henderson
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Richard Henderson
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
1999-03-31 23:46 ` Jeffrey A Law
1999-03-17 20:14 ` Jeffrey A Law
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
1999-03-31 23:46 ` Jeffrey A Law
1999-03-31 23:46 ` Jerry Quinn
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=18375.921819620@hurl.cygnus.com \
--to=law@hurl.cygnus.com \
--cc=egcs@egcs.cygnus.com \
--cc=jquinn@nortelnetworks.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).