public inbox for linux-fsdevel@vger.kernel.org
 help / color / mirror / Atom feed
From: Al Viro <viro@zeniv.linux.org.uk>
To: Jaroslav Kysela <perex@perex.cz>
Cc: linux-fsdevel@vger.kernel.org,
	"Amadeusz Sławiński" <amadeuszx.slawinski@linux.intel.com>,
	"Takashi Iwai" <tiwai@suse.de>,
	linux-sound@vger.kernel.org, "Vinod Koul" <vkoul@kernel.org>
Subject: Re: [CFT][PATCH] fix descriptor uses in sound/core/compress_offload.c
Date: Thu, 26 Dec 2024 22:17:26 +0000	[thread overview]
Message-ID: <20241226221726.GW1977892@ZenIV> (raw)
In-Reply-To: <20241226213122.GV1977892@ZenIV>

On Thu, Dec 26, 2024 at 09:31:22PM +0000, Al Viro wrote:
> On Thu, Dec 26, 2024 at 08:00:18PM +0100, Jaroslav Kysela wrote:
> 
> >   I already made almost similar patch:
> > 
> > https://lore.kernel.org/linux-sound/20241217100726.732863-1-perex@perex.cz/
> 
> Umm...  The only problem with your variant is that dma_buf_get()
> is wrong here - it should be get_dma_buf() on actual objects,
> and it should be done before fd_install().

Incremental on top of what just got merged into mainline:

Grab the references to dmabuf before moving them into descriptor
table - trying to do that by descriptor afterwards might end up getting
a different object, with a dangling reference left in task->{input,output}

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
diff --git a/sound/core/compress_offload.c b/sound/core/compress_offload.c
index edf5aadf38e5..543c7f525f84 100644
--- a/sound/core/compress_offload.c
+++ b/sound/core/compress_offload.c
@@ -1053,13 +1053,13 @@ static int snd_compr_task_new(struct snd_compr_stream *stream, struct snd_compr_
 		put_unused_fd(fd_i);
 		goto cleanup;
 	}
+	/* keep dmabuf reference until freed with task free ioctl */
+	get_dma_buf(task->input);
+	get_dma_buf(task->output);
 	fd_install(fd_i, task->input->file);
 	fd_install(fd_o, task->output->file);
 	utask->input_fd = fd_i;
 	utask->output_fd = fd_o;
-	/* keep dmabuf reference until freed with task free ioctl */
-	dma_buf_get(utask->input_fd);
-	dma_buf_get(utask->output_fd);
 	list_add_tail(&task->list, &stream->runtime->tasks);
 	stream->runtime->total_tasks++;
 	return 0;

  reply	other threads:[~2024-12-26 22:17 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-12-26 18:29 [CFT][PATCH] fix descriptor uses in sound/core/compress_offload.c Al Viro
2024-12-26 19:00 ` Jaroslav Kysela
2024-12-26 21:31   ` Al Viro
2024-12-26 22:17     ` Al Viro [this message]
2024-12-29  8:35       ` Takashi Iwai
2024-12-29 18:53         ` Al Viro

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=20241226221726.GW1977892@ZenIV \
    --to=viro@zeniv.linux.org.uk \
    --cc=amadeuszx.slawinski@linux.intel.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-sound@vger.kernel.org \
    --cc=perex@perex.cz \
    --cc=tiwai@suse.de \
    --cc=vkoul@kernel.org \
    /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