Hi! > From: Takashi Iwai <tiwai@xxxxxxx> > > [ Upstream commit 10457f5042b4890a667e2f15a2e783490dda44d2 ] > > So far the vmaster code has been tolerant about the input values and > accepts any values by correcting internally. But now our own selftest > starts complaining about this behavior, so let's be picky and change > the behavior to return -EINVAL for invalid input values instead. Does it fix a bug serious enough for -stable? Best regards, Pavel > +++ b/sound/core/vmaster.c > @@ -204,6 +204,12 @@ static int follower_put(struct snd_kcontrol *kcontrol, > err = follower_init(follower); > if (err < 0) > return err; > + for (ch = 0; ch < follower->info.count; ch++) { > + if (ucontrol->value.integer.value[ch] < follower->info.min_val || > + ucontrol->value.integer.value[ch] > follower->info.max_val) > + return -EINVAL; > + } > + > for (ch = 0; ch < follower->info.count; ch++) { > if (follower->vals[ch] != ucontrol->value.integer.value[ch]) { > changed = 1; > @@ -344,6 +350,8 @@ static int master_put(struct snd_kcontrol *kcontrol, > new_val = ucontrol->value.integer.value[0]; > if (new_val == old_val) > return 0; > + if (new_val < master->info.min_val || new_val > master->info.max_val) > + return -EINVAL; > > err = sync_followers(master, old_val, new_val); > if (err < 0) -- DENX Software Engineering GmbH, Managing Director: Erika Unter HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Attachment:
signature.asc
Description: PGP signature
- References:
- Prev by Date: [PATCH] ALSA: hda/realtek: Limit mic boost on Positivo DN50F
- Next by Date: [PATCH v2 03/17] ASoC: Intel: boards: always check the result of acpi_dev_get_first_match_dev()
- Previous by thread: [PATCH AUTOSEL 5.10 26/38] ALSA: vmaster: Return error for invalid input values
- Next by thread: [PATCH v24 00/34] Introduce QC USB SND audio offloading support
- Index(es):