netdev - Re: Calculate VIRTQUEUE_NUM in "net/9p/trans

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <ZxwTOB5ENi66C_kq@codewreck.org>
Date: Sat, 26 Oct 2024 06:52:56 +0900
From: Dominique Martinet <asmadeus@...ewreck.org>
To: Christian Schoenebeck <linux_oss@...debyte.com>,
	Guan Xin <guanx.bac@...il.com>
Cc: v9fs@...ts.linux.dev,
	Linux Kernel Network Developers <netdev@...r.kernel.org>,
	linux-fsdevel@...r.kernel.org,
	Eric Van Hensbergen <ericvh@...nel.org>
Subject: Re: Calculate VIRTQUEUE_NUM in "net/9p/trans_virtio.c" from stack
 size

Christian,

this is more up your alley, letting you comment as well as you weren't
even sent a copy in Ccs

Guan,

overall, please check Documentation/process/submitting-patches.rst -
this is missing [PATCH] in the mail header, missing some recipients that
you'd have gotten from get_maintiner.pl, and the commit title is a mess.

Have a look at other recent patches on https://lore.kernel.org/v9fs/

Guan Xin wrote on Sat, Oct 26, 2024 at 12:18:42AM +0800:
> For HPC applications the hard-coded VIRTQUEUE_NUM of 128 seems to
> limit the throughput of guest systems accessing cluster filesystems
> mounted on the host.
> 
> Just increase VIRTQUEUE_NUM for kernels with a
> larger stack.

You're replacing an hardcoded value with another, this could be made
dynamic e.g. as a module_param so someone could tune this based on their
actual needs (and test more easily); I'd more readily accept such a
patch.

> Author: GUAN Xin <guanx.bac@...il.com>

Author: tag doesn't exist and would be useless here as it's the mail you
sent the patch from.

> Signed-off-by: GUAN Xin <guanx.bac@...il.com>
> cc: Eric Van Hensbergen <ericvh@...nel.org>
> cc: v9fs@...ts.linux.dev
> cc: netdev@...r.kernel.org
> cc: linux-fsdevel@...r.kernel.org
> 
> --- net/9p/trans_virtio.c.orig  2024-10-25 10:25:09.390922517 +0800
> +++ net/9p/trans_virtio.c       2024-10-25 16:48:40.451680192 +0800
> @@ -31,11 +31,12 @@
> #include <net/9p/transport.h>
> #include <linux/scatterlist.h>
> #include <linux/swap.h>
> +#include <linux/thread_info.h>
> #include <linux/virtio.h>
> #include <linux/virtio_9p.h>
> #include "trans_common.h"
> 
> -#define VIRTQUEUE_NUM  128
> +#define VIRTQUEUE_NUM  (1 << (THREAD_SIZE_ORDER + PAGE_SHIFT - 6))

(FWIW that turned out to be 256 on my system)

> /* a single mutex to manage channel initialization and attachment */
> static DEFINE_MUTEX(virtio_9p_lock);
> 

-- 
Dominique Martinet | Asmadeus