mailing list of musl libc
 help / color / mirror / code / Atom feed
From: Dominic Chen <d.c.ddcc@gmail.com>
To: Rich Felker <dalias@libc.org>
Cc: fweimer@redhat.com, musl@lists.openwall.com
Subject: Re: [musl] Incorrect thread TID caching
Date: Fri, 12 Mar 2021 16:14:51 -0500	[thread overview]
Message-ID: <60a94ab1-4682-fd04-37ce-92d61207988d@gmail.com> (raw)
In-Reply-To: <20210217210748.GL11590@brightrain.aerifal.cx>

On 2/17/2021 4:07 PM, Rich Felker wrote:
>> Whether it's inlined isn't really a big deal; this is not a hot path.
>> It's more just a matter of how it needs to be split up at the source
>> level, and it seems to be messy whichever way we choose.
>>
>> Trying to avoid calling __clone doesn't seem like such a good idea,
>> since the child has to run on a new stack -- if we did avoid it we'd
>> need a new way to switch stacks. The generic __unmapself has a hack
>> to do this already that we could reuse without needing new
>> arch-specific glue though.
>>
>> I'll keep trying things and see if I come up with something not too
>> unreasonable.
>>
>> Attached is a draft of how clone() *could* work without refactoring
>> the pre/post logic from _Fork to use __clone. I don't particularly
>> like it, and CRTJMP abuse like this (which __unmapself also does, as
>> noted above) is not valid for FDPIC archs (it actually expects a code
>> address not a function pointer). I'll try a version the other way and
>> see how it looks.

Sorry for the delay. I did end up needing robust mutex functionality, 
and ended up using a variant of your patch to fix-up. There were two 
issues that I noticed:

 1. Prior to CRTJMP, the 'arg' argument needs to be moved into the first
    argument register for the local calling convention. I spent a little
    time trying to get an architecture-neutral version working using a
    wrapper function, but gave up and used some inline assembly to
    populate %rdi on x86_64.
 2. If a thread exits without unlocking a robust mutex, the program will
    subsequently SIGSEGV since the thread's memory region has been
    unmapped, but a pointer to the robust mutex remains on the internal
    linked list. Not sure where the fault lies here, but just thought
    I'd mention it.

Thanks,

Dominic


  reply	other threads:[~2021-03-12 21:15 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-03  4:04 Dominic Chen
2021-02-03  7:16 ` Florian Weimer
2021-02-03 19:21 ` Rich Felker
2021-02-03 20:21   ` Dominic Chen
2021-02-03 21:01     ` Rich Felker
2021-02-03 22:30       ` Dominic Chen
2021-02-03 22:55         ` Rich Felker
2021-02-15 16:56           ` Rich Felker
2021-02-17 19:49             ` Dominic Chen
2021-02-17 20:11               ` Rich Felker
2021-02-17 21:07                 ` Rich Felker
2021-03-12 21:14                   ` Dominic Chen [this message]
2021-02-04  3:28     ` Carlos O'Donell
2021-02-04  4:22       ` Dominic Chen
2021-02-04 16:15         ` Rich Felker

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=60a94ab1-4682-fd04-37ce-92d61207988d@gmail.com \
    --to=d.c.ddcc@gmail.com \
    --cc=dalias@libc.org \
    --cc=fweimer@redhat.com \
    --cc=musl@lists.openwall.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.
Code repositories for project(s) associated with this public inbox

	https://git.vuxu.org/mirror/musl/

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 NNTP newsgroup(s).