linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Calvin Owens <calvin@wbinvd.org>
To: David Sterba <dsterba@suse.cz>
Cc: Qu Wenruo <quwenruo.btrfs@gmx.com>,
	linux-kernel@vger.kernel.org, linux-btrfs@vger.kernel.org,
	Daniel Vacek <neelx@suse.com>, David Sterba <dsterba@suse.com>,
	Josef Bacik <josef@toxicpanda.com>, Chris Mason <clm@fb.com>
Subject: Re: [PATCH] btrfs: Accept and ignore compression level for lzo
Date: Fri, 22 Aug 2025 14:42:41 -0700	[thread overview]
Message-ID: <aKjkUSL1zxL0VNGC@mozart.vkv.me> (raw)
In-Reply-To: <20250822184547.GX22430@suse.cz>

On Friday 08/22 at 20:45 +0200, David Sterba wrote:
> On Fri, Aug 22, 2025 at 02:28:29AM -0700, Calvin Owens wrote:
> > On Friday 08/22 at 17:57 +0930, Qu Wenruo wrote:
> > > 在 2025/8/22 17:15, Calvin Owens 写道:
> > > > The compression level is meaningless for lzo, but before commit
> > > > 3f093ccb95f30 ("btrfs: harden parsing of compression mount options"),
> > > > it was silently ignored if passed.
> > > 
> > > Since LZO doesn't support compression level, why providing a level parameter
> > > in the first place?
> > 
> > Interpreting "no level" as "level is always one" doesn't seem that
> > unreasonable to me, especially since it has worked forever.
> 
> As it currently works, no level means use the default, which is defined
> for all compression. For LZO it's implicit and 1.
> 
> > > I think it's time for those users to properly update their mount options.
> > 
> > It's a user visable regression, and fixing it has zero possible
> > downside. I think you should take my patch :)
> 
> I tend to agree this is a usability regression, even if LZO is a bit odd
> with levels, accepting the allowed values should work.
> 
> The mount options and level combinations that should work:
> 
> - compress=NAME   - use default level for NAME
> - compress=NAME:0 - use default, while accepting the level setting
> - compress=NAME:N - if N is in the allowed range for NAME then take it
> 
> The syntax is consistent for all three compressions.

Thanks David.

Maybe the below is a little more palatable? Letting the single level be
a detail so the branches in btrfs_parse_compress() all match?

But, the compression level ends up being printk'd as '1', where it has
always been '0' in the past (and still is in 6.17-rc):

    - BTRFS info (device vda state M): use lzo compression, level 0
    + BTRFS info (device vda state M): use lzo compression, level 1

With my v1 it's still always printed as zero, if that's preferable.

-----8<-----
From: Calvin Owens <calvin@wbinvd.org>
Subject: [PATCH v2] btrfs: Accept and ignore compression level for lzo

The compression level is meaningless for lzo, but before commit
3f093ccb95f30 ("btrfs: harden parsing of compression mount options"),
it was silently ignored if passed.

After that commit, passing a level with lzo fails to mount:

    BTRFS error: unrecognized compression value lzo:1

Restore the old behavior, in case any users were relying on it.

Fixes: 3f093ccb95f30 ("btrfs: harden parsing of compression mount options")
Signed-off-by: Calvin Owens <calvin@wbinvd.org>
---
 fs/btrfs/super.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index a262b494a89f..bbcaac7022b0 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -299,9 +299,10 @@ static int btrfs_parse_compress(struct btrfs_fs_context *ctx,
 		btrfs_set_opt(ctx->mount_opt, COMPRESS);
 		btrfs_clear_opt(ctx->mount_opt, NODATACOW);
 		btrfs_clear_opt(ctx->mount_opt, NODATASUM);
-	} else if (btrfs_match_compress_type(string, "lzo", false)) {
+	} else if (btrfs_match_compress_type(string, "lzo", true)) {
 		ctx->compress_type = BTRFS_COMPRESS_LZO;
-		ctx->compress_level = 0;
+		ctx->compress_level = btrfs_compress_str2level(BTRFS_COMPRESS_LZO,
+							       string + 4);
 		btrfs_set_opt(ctx->mount_opt, COMPRESS);
 		btrfs_clear_opt(ctx->mount_opt, NODATACOW);
 		btrfs_clear_opt(ctx->mount_opt, NODATASUM);
-- 
2.47.2


  reply	other threads:[~2025-08-22 21:42 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-22  7:45 [PATCH] btrfs: Accept and ignore compression level for lzo Calvin Owens
2025-08-22  8:27 ` Qu Wenruo
2025-08-22  9:28   ` Calvin Owens
2025-08-22 10:18     ` Qu Wenruo
2025-08-22 18:45     ` David Sterba
2025-08-22 21:42       ` Calvin Owens [this message]
2025-08-22 10:20 ` Sun YangKai
2025-08-22 10:23   ` Qu Wenruo
2025-08-22 15:54     ` Calvin Owens
2025-08-22 18:57       ` David Sterba
2025-08-22 21:44       ` Qu Wenruo
2025-08-22 23:24         ` Calvin Owens
2025-08-22 23:39           ` Qu Wenruo
2025-08-24 15:58             ` Calvin Owens
2025-08-24 21:40               ` Qu Wenruo
2025-08-25  8:51               ` Daniel Vacek
2025-08-25  9:03                 ` Qu Wenruo
2025-08-25 11:37                   ` Filipe Manana
2025-08-22 18:42   ` Calvin Owens

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=aKjkUSL1zxL0VNGC@mozart.vkv.me \
    --to=calvin@wbinvd.org \
    --cc=clm@fb.com \
    --cc=dsterba@suse.com \
    --cc=dsterba@suse.cz \
    --cc=josef@toxicpanda.com \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=neelx@suse.com \
    --cc=quwenruo.btrfs@gmx.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).