Linux-EROFS Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Gao Xiang <xiang@kernel.org>
To: Zhan Xusheng <zhanxusheng1024@gmail.com>
Cc: Gao Xiang <xiang@kernel.org>,
	Jingbo Xu <jefflexu@linux.alibaba.com>,
	zhanxusheng@xiaomi.com, linux-erofs@lists.ozlabs.org
Subject: Re: [PATCH] erofs-utils: tests: add a splice() test for inode_share mode
Date: Fri, 21 Aug 2026 14:36:17 +0800	[thread overview]
Message-ID: <aofx4XAjsCKqo9k_@XiangdeMacBook-Pro.local> (raw)
In-Reply-To: <20260821015722.1936404-1-zhanxusheng@xiaomi.com>

Hi Xusheng,

On Fri, Aug 21, 2026 at 09:57:22AM +0800, Zhan Xusheng wrote:
> From: Zhan Xusheng <zhanxusheng1024@gmail.com>
> 
> From: Zhan Xusheng <zhanxusheng@xiaomi.com>
> 
> splice() and sendfile() went through filemap_splice_read() on the user
> file, so they populated the per-inode page cache instead of the shared one
> that read() and mmap() use.  The content was correct either way, which is
> why nothing caught it: only which mapping the pages land in changes.
> 
> Observe that instead.  cachestat() works on the file's own mapping, while
> fincore's mmap()+mincore() fallback goes through erofs_ishare_mmap(), which

Actually we shouldn't rely on the fact that cachestat() works on the
file's own mapping since it's another misfeature, similar to:

https://lore.kernel.org/linux-unionfs/20260623111533.2285005-1-ptikhomirov@virtuozzo.com/

we should find another way to replace the cachestat() usage here.

Thanks,
Gao Xiang


      reply	other threads:[~2026-08-21  6:36 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-21  1:57 [PATCH] erofs-utils: tests: add a splice() test for inode_share mode Zhan Xusheng
2026-08-21  6:36 ` Gao Xiang [this message]

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=aofx4XAjsCKqo9k_@XiangdeMacBook-Pro.local \
    --to=xiang@kernel.org \
    --cc=jefflexu@linux.alibaba.com \
    --cc=linux-erofs@lists.ozlabs.org \
    --cc=zhanxusheng1024@gmail.com \
    --cc=zhanxusheng@xiaomi.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