From: "Martin Liška" <mliska@suse.cz>
To: Jan Hubicka <hubicka@ucw.cz>, Jakub Jelinek <jakub@redhat.com>
Cc: Jeff Law <law@redhat.com>,
gcc-patches@gcc.gnu.org, Michael Matz <matz@suse.de>,
Richard Biener <richard.guenther@gmail.com>
Subject: Re: [PATCH] Deduce automatically number of cores for -flto option.
Date: Wed, 31 Jul 2019 07:50:00 -0000 [thread overview]
Message-ID: <217ef7a3-7726-04fa-3bd5-ca5b10844091@suse.cz> (raw)
In-Reply-To: <20190731074019.b4l6pwcpeh2q2zai@kam.mff.cuni.cz>
On 7/31/19 9:40 AM, Jan Hubicka wrote:
>> Neither will work very well, we have thousands of -flto tests outside of
>> lto.exp, e.g. dg-torture.exp (or libgomp and others) cycle through various
>> options including -flto etc.
>>
>> Some env var would be useful I guess, though shouldn't it have GCC in the
>> name? I mean, if we run into these fork-bomb problems in gcc, won't other
>> projects run into those as well?
>>
>> Why doesn't the jobserver work in the tests? Is that because of missing +
>> somewhere in the Makefiles or is something unsetting MFLAGS or MAKEFLAGS
>> env vars?
>
> Main trouble with make's jobserver is that it works by
> 1) defining environment variable saying which file descriptior to
> connect to
> 2) keeping the file descriptor open upon invoking "+" prefixed lines
>
> Adding "+" to GCC invocation is wrong since it breaks dry run (we do not
> want to link at that time) but it is only way to access the jobserver.
> If "+" is not present, make will keep the environment vairable but will
> close file descriptors prior exec.
>
> Make developers said that this is because some old prorams misbehave
> when you exec them with more than 3 file descriptors open. I tried to
> negotiate for named pipe which would solve this and it would make it
> easy to connect to outermost jobserver from anything invoked form
> toplevel make, but they was worried about systems w/o named pipes.
Yes a more generic approach would be welcome as other build systems
would also be able to utilize it.
>
> I wonder why we do not detect jobserv as unavailable in this case and do
> not default to -flto=<numthreads>?
We do not detect jobserver because of Dejagnu is not using it.
And yes, we default to -flto=<numthreads> in LTO tests now. That's
what is causing issues right now.
> Is it because dejagnu machinery actually opens some other file
> descriptor that gets same ID and executes us with it?
>
> Or does LTO wrapper open something prior accessing jobserver?
>
>>
>> Though, as I said on IRC, I think we might run out of filedescriptors when
>> using jobserver too, say if on 64 thread machine one does make -j64 -k check
>> and each test simultaneously tries to create 64 partitions, that would be
>> 4096 connections to the jobserver, right?
>
> Only WPA process connects to jobserver (which is 1 per linker
> invokation), so I think this should be safe.
Yes and it's only about a quick fcntl. But as mentioned, Dejagnu is not passing
us jobserver, so we don't do it right now.
Martin
>
> Honza
>>
>> Jakub
next prev parent reply other threads:[~2019-07-31 7:49 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-23 8:55 [PATCH] Come up with -flto=auto option Martin Liška
2019-07-23 9:29 ` Jan Hubicka
2019-07-23 10:34 ` [PATCH] Deduce automatically number of cores for -flto option Martin Liška
2019-07-24 15:47 ` Jeff Law
2019-07-29 13:37 ` Martin Liška
2019-07-30 13:47 ` Martin Liška
2019-07-31 1:23 ` Jakub Jelinek
2019-07-31 7:24 ` Martin Liška
2019-07-31 7:40 ` Jakub Jelinek
2019-07-31 7:49 ` Jan Hubicka
2019-07-31 7:50 ` Martin Liška [this message]
2019-07-31 7:54 ` Martin Liška
2019-07-31 8:08 ` Jakub Jelinek
2019-07-31 8:21 ` Jan Hubicka
2019-07-31 8:37 ` Martin Liška
2019-07-31 9:12 ` Jakub Jelinek
2019-07-31 9:15 ` Jan Hubicka
2019-07-31 9:17 ` Jakub Jelinek
2019-07-31 9:22 ` Jan Hubicka
2019-07-31 10:02 ` Martin Liška
2019-07-31 12:02 ` Jan Hubicka
2019-07-31 15:42 ` Martin Liška
2019-08-01 13:19 ` Jakub Jelinek
2019-08-01 14:34 ` [PATCH] Properly detect working jobserver in gcc driver Martin Liška
2019-08-01 14:41 ` Jakub Jelinek
2019-08-02 6:30 ` Martin Liška
2019-08-02 7:45 ` Jakub Jelinek
2019-08-02 8:47 ` Martin Liška
2019-08-02 8:50 ` Jakub Jelinek
2019-08-02 9:04 ` Richard Biener
2019-08-02 9:08 ` Jan Hubicka
2019-08-02 9:15 ` Jan Hubicka
2019-08-02 9:19 ` Martin Liška
2019-08-02 9:55 ` Richard Biener
2019-08-05 6:41 ` Martin Liška
2019-08-09 8:14 ` Martin Liška
2019-08-09 8:22 ` Richard Biener
2019-08-09 12:51 ` Martin Liška
2019-08-09 13:56 ` Martin Liška
2019-08-12 15:18 ` Jeff Law
2019-07-23 13:13 ` [PATCH] Come up with -flto=auto option Jeff Law
2019-07-23 13:22 ` Richard Biener
2019-07-23 13:57 ` Michael Matz
2019-07-23 14:00 ` Jeff Law
2019-07-23 14:27 ` Martin Liška
2019-07-23 22:56 ` Jeff Law
2019-07-24 6:59 ` Martin Liška
2019-07-24 15:16 ` Jeff Law
2019-07-23 22:32 ` Allan Sandfeld Jensen
2019-07-24 6:47 ` Martin Liška
2019-07-24 7:12 ` Allan Sandfeld Jensen
2019-07-24 7:15 ` Martin Liška
2019-07-24 11:09 ` Nathan Sidwell
2019-07-24 15:46 ` Jeff Law
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=217ef7a3-7726-04fa-3bd5-ca5b10844091@suse.cz \
--to=mliska@suse.cz \
--cc=gcc-patches@gcc.gnu.org \
--cc=hubicka@ucw.cz \
--cc=jakub@redhat.com \
--cc=law@redhat.com \
--cc=matz@suse.de \
--cc=richard.guenther@gmail.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).