Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1668845 > unrolled thread
| Started by | Robert Jarzmik <robert.jarzmik@free.fr> |
|---|---|
| First post | 2017-06-19 09:30 +0200 |
| Last post | 2017-06-19 14:00 +0200 |
| Articles | 3 — 2 participants |
Back to article view | Back to linux.kernel
This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by
below is the oldest one visible, not the original post.
[PATCH v2 12/12] ASoC: Fix use-after-free at card unregistration Robert Jarzmik <robert.jarzmik@free.fr> - 2017-06-19 09:30 +0200
Re: [PATCH v2 12/12] ASoC: Fix use-after-free at card unregistration Takashi Iwai <tiwai@suse.de> - 2017-06-19 11:30 +0200
Re: [PATCH v2 12/12] ASoC: Fix use-after-free at card unregistration Robert Jarzmik <robert.jarzmik@free.fr> - 2017-06-19 14:00 +0200
| From | Robert Jarzmik <robert.jarzmik@free.fr> |
|---|---|
| Date | 2017-06-19 09:30 +0200 |
| Subject | [PATCH v2 12/12] ASoC: Fix use-after-free at card unregistration |
| Message-ID | <tTZih-83v-3@gated-at.bofh.it> |
From: Takashi Iwai <tiwai@suse.de> soc_cleanup_card_resources() call snd_card_free() at the last of its procedure. This turned out to lead to a use-after-free. PCM runtimes have been already removed via soc_remove_pcm_runtimes(), while it's dereferenced later in soc_pcm_free() called via snd_card_free(). The fix is simple: just move the snd_card_free() call to the beginning of the whole procedure. This also gives another benefit: it guarantees that all operations have been shut down before actually releasing the resources, which was racy until now. Reported-and-tested-by: Robert Jarzmik <robert.jarzmik@free.fr> Cc: <stable@vger.kernel.org> Signed-off-by: Takashi Iwai <tiwai@suse.de> --- sound/soc/soc-core.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index 2722bb0c5573..98d60f471c5d 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -2286,6 +2286,9 @@ static int soc_cleanup_card_resources(struct snd_soc_card *card) list_for_each_entry(rtd, &card->rtd_list, list) flush_delayed_work(&rtd->delayed_work); + /* free the ALSA card at first; this syncs with pending operations */ + snd_card_free(card->snd_card); + /* remove and free each DAI */ soc_remove_dai_links(card); soc_remove_pcm_runtimes(card); @@ -2300,9 +2303,7 @@ static int soc_cleanup_card_resources(struct snd_soc_card *card) if (card->remove) card->remove(card); - snd_card_free(card->snd_card); return 0; - } /* removes a socdev */ -- 2.1.4
[toc] | [next] | [standalone]
| From | Takashi Iwai <tiwai@suse.de> |
|---|---|
| Date | 2017-06-19 11:30 +0200 |
| Message-ID | <tU1aq-ON-11@gated-at.bofh.it> |
| In reply to | #1668845 |
On Mon, 19 Jun 2017 09:27:09 +0200, Robert Jarzmik wrote: > > From: Takashi Iwai <tiwai@suse.de> > > soc_cleanup_card_resources() call snd_card_free() at the last of its > procedure. This turned out to lead to a use-after-free. > PCM runtimes have been already removed via soc_remove_pcm_runtimes(), > while it's dereferenced later in soc_pcm_free() called via > snd_card_free(). > > The fix is simple: just move the snd_card_free() call to the beginning > of the whole procedure. This also gives another benefit: it > guarantees that all operations have been shut down before actually > releasing the resources, which was racy until now. > > Reported-and-tested-by: Robert Jarzmik <robert.jarzmik@free.fr> > Cc: <stable@vger.kernel.org> > Signed-off-by: Takashi Iwai <tiwai@suse.de> This patch must be superfluous :) Takashi > --- > sound/soc/soc-core.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c > index 2722bb0c5573..98d60f471c5d 100644 > --- a/sound/soc/soc-core.c > +++ b/sound/soc/soc-core.c > @@ -2286,6 +2286,9 @@ static int soc_cleanup_card_resources(struct snd_soc_card *card) > list_for_each_entry(rtd, &card->rtd_list, list) > flush_delayed_work(&rtd->delayed_work); > > + /* free the ALSA card at first; this syncs with pending operations */ > + snd_card_free(card->snd_card); > + > /* remove and free each DAI */ > soc_remove_dai_links(card); > soc_remove_pcm_runtimes(card); > @@ -2300,9 +2303,7 @@ static int soc_cleanup_card_resources(struct snd_soc_card *card) > if (card->remove) > card->remove(card); > > - snd_card_free(card->snd_card); > return 0; > - > } > > /* removes a socdev */ > -- > 2.1.4 > >
[toc] | [prev] | [next] | [standalone]
| From | Robert Jarzmik <robert.jarzmik@free.fr> |
|---|---|
| Date | 2017-06-19 14:00 +0200 |
| Message-ID | <tU3vA-29m-1@gated-at.bofh.it> |
| In reply to | #1668931 |
Takashi Iwai <tiwai@suse.de> writes: > On Mon, 19 Jun 2017 09:27:09 +0200, > Robert Jarzmik wrote: >> >> From: Takashi Iwai <tiwai@suse.de> >> >> soc_cleanup_card_resources() call snd_card_free() at the last of its >> procedure. This turned out to lead to a use-after-free. >> PCM runtimes have been already removed via soc_remove_pcm_runtimes(), >> while it's dereferenced later in soc_pcm_free() called via >> snd_card_free(). >> >> The fix is simple: just move the snd_card_free() call to the beginning >> of the whole procedure. This also gives another benefit: it >> guarantees that all operations have been shut down before actually >> releasing the resources, which was racy until now. >> >> Reported-and-tested-by: Robert Jarzmik <robert.jarzmik@free.fr> >> Cc: <stable@vger.kernel.org> >> Signed-off-by: Takashi Iwai <tiwai@suse.de> > > This patch must be superfluous :) Haha :) My serie shifted by one, so the very first of the serie is therefore missing, formerly "ALSA: ac97: split out the generic ac97 registers" in https://patchwork.kernel.org/patch/9398143/, and the shift triggered the inclusion of the last patch of my tree, ie. yours :) Cheers. -- Robert
[toc] | [prev] | [standalone]
Back to top | Article view | linux.kernel
csiph-web