From: Olivier Hainque <hainque@adacore.com>
To: Segher Boessenkool <segher@kernel.crashing.org>
Cc: Olivier Hainque <hainque@adacore.com>,
GCC Patches <gcc-patches@gcc.gnu.org>,
Alan Modra <amodra@gmail.com>
Subject: Re: rs6000 stack_tie mishap again
Date: Wed, 30 Mar 2016 09:40:00 -0000 [thread overview]
Message-ID: <891E961D-CC21-45ED-AEBD-1DB16922DBEC@adacore.com> (raw)
In-Reply-To: <20160328111812.GA17381@gate.crashing.org>
Hello Segher,
> On Mar 28, 2016, at 13:18 , Segher Boessenkool <segher@kernel.crashing.org> wrote:
>
>> You need to have had r11 last used to designate a global
>> symbol as part of the function body (in order to have base_term
>> designate a symbol_ref etc), and then have the scheduler
>> decide that moving across is a good idea. It's certainly not
>> an easy combination to trigger.
>
> Yes, I did that (with some asm's). Like this:
>
> ===
> void g(int, char *);
>
> int dum;
>
> void f(int x)
> {
> char big[200000];
> g(x, big);
> g(x, big);
> register void *p asm("r11") = &dum;
> asm("" : : "r"(p));
> }
Ah, I see, thanks. In this instance, the problem doesn't
trigger because CONSTANT_POOL_ADDRESS_P (base) is false in
base = find_base_term (true_mem_addr);
if (! writep
&& base
&& (GET_CODE (base) == LABEL_REF
|| (GET_CODE (base) == SYMBOL_REF
&& CONSTANT_POOL_ADDRESS_P (base))))
return 0;
(part of write_dependence_p)
With a minor variation:
void g(int, char *);
void f(int x)
{
char big[200000];
start:
g(x, big);
g(x, big);
register void *p asm("r11") = &&start;
asm("" : : "r"(p));
asm("" : : :"r28");
asm("" : : :"r29");
asm("" : : :"r30");
}
I'm getting:
lis 11,.L2@ha
la 11,.L2@l(11)
lwz 11,0(1)
lwz 0,4(11)
lwz 28,-16(11)
mr 1,11
mtlr 0
lwz 29,-12(11)
lwz 30,-8(11)
lwz 31,-4(11)
blr
out of a powerpc-elf close-to-tunk compiler, despite the
presence of a stack_tie insn at the rtl level.
Olivier
next prev parent reply other threads:[~2016-03-30 9:02 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-23 16:24 Olivier Hainque
2016-03-24 7:51 ` Alan Modra
2016-03-24 10:32 ` Olivier Hainque
2016-03-24 17:00 ` Jeff Law
2016-03-28 19:58 ` Richard Henderson
2016-04-08 8:25 ` Olivier Hainque
2016-04-08 15:37 ` Segher Boessenkool
2016-04-08 16:01 ` Olivier Hainque
2016-04-11 10:15 ` Olivier Hainque
2016-04-14 15:47 ` Andreas Krebbel
2016-04-14 16:50 ` Jeff Law
2016-04-14 17:10 ` Olivier Hainque
2016-04-15 4:37 ` Andreas Krebbel
2016-04-15 7:43 ` Olivier Hainque
2016-04-15 16:42 ` Jeff Law
2016-04-15 17:05 ` Olivier Hainque
2016-04-15 17:26 ` Segher Boessenkool
2016-04-14 22:42 ` Segher Boessenkool
2016-04-15 15:17 ` Olivier Hainque
2016-03-28 4:48 ` Segher Boessenkool
2016-03-28 11:23 ` Olivier Hainque
2016-03-28 12:44 ` Segher Boessenkool
2016-03-30 9:40 ` Olivier Hainque [this message]
2016-03-30 15:15 ` Alan Modra
2016-03-23 17:42 David Edelsohn
2016-03-24 8:17 ` Alan Modra
2016-03-24 10:17 ` Olivier Hainque
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=891E961D-CC21-45ED-AEBD-1DB16922DBEC@adacore.com \
--to=hainque@adacore.com \
--cc=amodra@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).