qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Daniel P. Berrangé" <berrange@redhat.com>
To: Stefan Hajnoczi <stefanha@redhat.com>
Cc: qemu-devel@nongnu.org, Kevin Wolf <kwolf@redhat.com>,
	Sanjay Rao <srao@redhat.com>,
	Boaz Ben Shabat <bbenshab@redhat.com>,
	Joe Mario <jmario@redhat.com>
Subject: Re: [PATCH] coroutine: cap per-thread local pool size
Date: Tue, 19 Mar 2024 13:43:32 +0000	[thread overview]
Message-ID: <ZfmWhDaG5mN-GCeO@redhat.com> (raw)
In-Reply-To: <20240318183429.1039340-1-stefanha@redhat.com>

On Mon, Mar 18, 2024 at 02:34:29PM -0400, Stefan Hajnoczi wrote:
> The coroutine pool implementation can hit the Linux vm.max_map_count
> limit, causing QEMU to abort with "failed to allocate memory for stack"
> or "failed to set up stack guard page" during coroutine creation.
> 
> This happens because per-thread pools can grow to tens of thousands of
> coroutines. Each coroutine causes 2 virtual memory areas to be created.

This sounds quite alarming. What usage scenario is justified in
creating so many coroutines ?

IIUC, coroutine stack size is 1 MB, and so tens of thousands of
coroutines implies 10's of GB of memory just on stacks alone.

> Eventually vm.max_map_count is reached and memory-related syscalls fail.

On my system max_map_count is 1048576, quite alot higher than
10's of 1000's. Hitting that would imply ~500,000 coroutines and
~500 GB of stacks !

> diff --git a/util/qemu-coroutine.c b/util/qemu-coroutine.c
> index 5fd2dbaf8b..2790959eaf 100644
> --- a/util/qemu-coroutine.c
> +++ b/util/qemu-coroutine.c

> +static unsigned int get_global_pool_hard_max_size(void)
> +{
> +#ifdef __linux__
> +    g_autofree char *contents = NULL;
> +    int max_map_count;
> +
> +    /*
> +     * Linux processes can have up to max_map_count virtual memory areas
> +     * (VMAs). mmap(2), mprotect(2), etc fail with ENOMEM beyond this limit. We
> +     * must limit the coroutine pool to a safe size to avoid running out of
> +     * VMAs.
> +     */
> +    if (g_file_get_contents("/proc/sys/vm/max_map_count", &contents, NULL,
> +                            NULL) &&
> +        qemu_strtoi(contents, NULL, 10, &max_map_count) == 0) {
> +        /*
> +         * This is a conservative upper bound that avoids exceeding
> +         * max_map_count. Leave half for non-coroutine users like library
> +         * dependencies, vhost-user, etc. Each coroutine takes up 2 VMAs so
> +         * halve the amount again.
> +         */
> +        return max_map_count / 4;

That's 256,000 coroutines, which still sounds incredibly large
to me.

> +    }
> +#endif
> +
> +    return UINT_MAX;

Why UINT_MAX as a default ?  If we can't read procfs, we should
assume some much smaller sane default IMHO, that corresponds to
what current linux default max_map_count would be.

> +}
> +
> +static void __attribute__((constructor)) qemu_coroutine_init(void)
> +{
> +    qemu_mutex_init(&global_pool_lock);
> +    global_pool_hard_max_size = get_global_pool_hard_max_size();
>  }
> -- 
> 2.44.0
> 
> 

With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|



  parent reply	other threads:[~2024-03-19 13:44 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-18 18:34 [PATCH] coroutine: cap per-thread local pool size Stefan Hajnoczi
2024-03-19 13:32 ` Kevin Wolf
2024-03-19 13:45   ` Stefan Hajnoczi
2024-03-19 14:23   ` Sanjay Rao
2024-03-19 13:43 ` Daniel P. Berrangé [this message]
2024-03-19 16:54   ` Kevin Wolf
2024-03-19 17:10     ` Daniel P. Berrangé
2024-03-19 17:41       ` Kevin Wolf
2024-03-19 20:14         ` Daniel P. Berrangé
2024-03-19 17:55   ` Stefan Hajnoczi
2024-03-19 20:10     ` Daniel P. Berrangé
2024-03-20 13:35       ` Stefan Hajnoczi
2024-03-20 14:09         ` Daniel P. Berrangé
2024-03-21 12:21           ` Kevin Wolf
2024-03-21 16:59             ` Stefan Hajnoczi

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=ZfmWhDaG5mN-GCeO@redhat.com \
    --to=berrange@redhat.com \
    --cc=bbenshab@redhat.com \
    --cc=jmario@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=srao@redhat.com \
    --cc=stefanha@redhat.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 NNTP newsgroup(s).