December 2021 - Alsa-devel - mailman.alsa-project.org

[PATCH v2] ALSA: gus: Fix memory leaks at memory allocator error paths
by Takashi Iwai 13 Dec '21

13 Dec '21

When snd_gf1_mem_xalloc() returns NULL, the current code still leaves the formerly allocated block.name string but returns an error immediately. This patch does code-refactoring to move the kstrdup() call itself into snd_gf1_mem_xalloc() and deals with the resource free in the helper code by itself for fixing those memory leaks. Suggested-by: Jaroslav Kysela <perex(a)perex.cz> Link: https://lore.kernel.org/r/20211213132444.22385-2-tiwai@suse.de Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/isa/gus/gus_mem.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/sound/isa/gus/gus_mem.c b/sound/isa/gus/gus_mem.c index 5e3ff3137dd7..3e56c01c4544 100644 --- a/sound/isa/gus/gus_mem.c +++ b/sound/isa/gus/gus_mem.c @@ -24,8 +24,9 @@ void snd_gf1_mem_lock(struct snd_gf1_mem * alloc, int xup) } } -static struct snd_gf1_mem_block *snd_gf1_mem_xalloc(struct snd_gf1_mem * alloc, - struct snd_gf1_mem_block * block) +static struct snd_gf1_mem_block * +snd_gf1_mem_xalloc(struct snd_gf1_mem *alloc, struct snd_gf1_mem_block *block, + const char *name) { struct snd_gf1_mem_block *pblock, *nblock; @@ -33,6 +34,12 @@ static struct snd_gf1_mem_block *snd_gf1_mem_xalloc(struct snd_gf1_mem * alloc, if (nblock == NULL) return NULL; *nblock = *block; + nblock->name = kstrdup(name, GFP_KERNEL); + if (!nblock->name) { + kfree(nblock); + return NULL; + } + pblock = alloc->first; while (pblock) { if (pblock->ptr > nblock->ptr) { @@ -198,12 +205,7 @@ struct snd_gf1_mem_block *snd_gf1_mem_alloc(struct snd_gf1_mem * alloc, int owne if (share_id != NULL) memcpy(&block.share_id, share_id, sizeof(block.share_id)); block.owner = owner; - block.name = kstrdup(name, GFP_KERNEL); - if (block.name == NULL) { - snd_gf1_mem_lock(alloc, 1); - return NULL; - } - nblock = snd_gf1_mem_xalloc(alloc, &block); + nblock = snd_gf1_mem_xalloc(alloc, &block, name); snd_gf1_mem_lock(alloc, 1); return nblock; } @@ -240,14 +242,12 @@ int snd_gf1_mem_init(struct snd_gus_card * gus) if (gus->gf1.enh_mode) { block.ptr = 0; block.size = 1024; - block.name = kstrdup("InterWave LFOs", GFP_KERNEL); - if (block.name == NULL || snd_gf1_mem_xalloc(alloc, &block) == NULL) + if (!snd_gf1_mem_xalloc(alloc, &block, "InterWave LFOs")) return -ENOMEM; } block.ptr = gus->gf1.default_voice_address; block.size = 4; - block.name = kstrdup("Voice default (NULL's)", GFP_KERNEL); - if (block.name == NULL || snd_gf1_mem_xalloc(alloc, &block) == NULL) + if (!snd_gf1_mem_xalloc(alloc, &block, "Voice default (NULL's)")) return -ENOMEM; #ifdef CONFIG_SND_DEBUG snd_card_ro_proc_new(gus->card, "gusmem", gus, snd_gf1_mem_info_read); -- 2.31.1

2 1

alsa crashing each time my screen gets locked
by GitHub issues - opened 13 Dec '21

13 Dec '21

alsa-project/alsa-lib issue #203 was opened from Fallast: each time i get my screen lock after being inactive for 5 minutes i log back in and alsa stops working and my volume icon disappears i have to run volumeicon to get it working back again i'm on mx linux, based on debian bullseye Issue URL : https://github.com/alsa-project/alsa-lib/issues/203 Repository URL: https://github.com/alsa-project/alsa-lib

1 0

[PATCH 1/2] ALSA: gus: Fix erroneous memory allocation
by Takashi Iwai 13 Dec '21

13 Dec '21

snd_gf1_mem_xalloc() returns NULL incorrectly when the memory chunk is allocated in the middle of the chain. This patch corrects the return value to treat it properly. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/isa/gus/gus_mem.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/isa/gus/gus_mem.c b/sound/isa/gus/gus_mem.c index 4c691dbf2721..5e3ff3137dd7 100644 --- a/sound/isa/gus/gus_mem.c +++ b/sound/isa/gus/gus_mem.c @@ -44,7 +44,7 @@ static struct snd_gf1_mem_block *snd_gf1_mem_xalloc(struct snd_gf1_mem * alloc, else nblock->prev->next = nblock; mutex_unlock(&alloc->memory_mutex); - return NULL; + return nblock; } pblock = pblock->next; } -- 2.31.1

2 4

[PATCH v2] ALSA: sound/isa/gus: check the return value of kstrdup()
by Xiaoke Wang 13 Dec '21

13 Dec '21

kstrdup() returns NULL when some internal memory errors happen, it is better to check the return value of it. Otherwise, we may not to be able to catch some memory errors in time. Signed-off-by: Xiaoke Wang <xkernel.wang(a)foxmail.com> --- Changes in v2: - add the proper error handling. Thanks for the suggestions from Takashi Iwai. In the future, I'll pay more attention to what he mentioned. --- sound/isa/gus/gus_mem.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/sound/isa/gus/gus_mem.c b/sound/isa/gus/gus_mem.c index ff9480f..4c691db 100644 --- a/sound/isa/gus/gus_mem.c +++ b/sound/isa/gus/gus_mem.c @@ -199,6 +199,10 @@ struct snd_gf1_mem_block *snd_gf1_mem_alloc(struct snd_gf1_mem * alloc, int owne memcpy(&block.share_id, share_id, sizeof(block.share_id)); block.owner = owner; block.name = kstrdup(name, GFP_KERNEL); + if (block.name == NULL) { + snd_gf1_mem_lock(alloc, 1); + return NULL; + } nblock = snd_gf1_mem_xalloc(alloc, &block); snd_gf1_mem_lock(alloc, 1); return nblock; @@ -237,13 +241,13 @@ int snd_gf1_mem_init(struct snd_gus_card * gus) block.ptr = 0; block.size = 1024; block.name = kstrdup("InterWave LFOs", GFP_KERNEL); - if (snd_gf1_mem_xalloc(alloc, &block) == NULL) + if (block.name == NULL || snd_gf1_mem_xalloc(alloc, &block) == NULL) return -ENOMEM; } block.ptr = gus->gf1.default_voice_address; block.size = 4; block.name = kstrdup("Voice default (NULL's)", GFP_KERNEL); - if (snd_gf1_mem_xalloc(alloc, &block) == NULL) + if (block.name == NULL || snd_gf1_mem_xalloc(alloc, &block) == NULL) return -ENOMEM; #ifdef CONFIG_SND_DEBUG snd_card_ro_proc_new(gus->card, "gusmem", gus, snd_gf1_mem_info_read); --

2 1

Re: [PATCH] ALSA: sound/isa/gus: check the return value of kstrdup()
by Xiaoke Wang 13 Dec '21

13 Dec '21

Note: Compare with the last email, this one is using my full name. kstrdup() returns NULL when some internal memory errors happen, it is better to check the return value of it. Otherwise, we may not to be able to catch some memory errors in time. Signed-off-by: Xiaoke Wang <xkernel.wang(a)foxmail.com> --- sound/isa/gus/gus_mem.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/sound/isa/gus/gus_mem.c b/sound/isa/gus/gus_mem.c index ff9480f..f8d915f 100644 --- a/sound/isa/gus/gus_mem.c +++ b/sound/isa/gus/gus_mem.c @@ -199,6 +199,8 @@ struct snd_gf1_mem_block *snd_gf1_mem_alloc(struct snd_gf1_mem * alloc, int owne memcpy(&block.share_id, share_id, sizeof(block.share_id)); block.owner = owner; block.name = kstrdup(name, GFP_KERNEL); + if (block.name == NULL) + return NULL; nblock = snd_gf1_mem_xalloc(alloc, &block); snd_gf1_mem_lock(alloc, 1); return nblock; @@ -237,13 +239,13 @@ int snd_gf1_mem_init(struct snd_gus_card * gus) block.ptr = 0; block.size = 1024; block.name = kstrdup("InterWave LFOs", GFP_KERNEL); - if (snd_gf1_mem_xalloc(alloc, &block) == NULL) + if (block.name == NULL || snd_gf1_mem_xalloc(alloc, &block) == NULL) return -ENOMEM; } block.ptr = gus->gf1.default_voice_address; block.size = 4; block.name = kstrdup("Voice default (NULL's)", GFP_KERNEL); - if (snd_gf1_mem_xalloc(alloc, &block) == NULL) + if (block.name == NULL || snd_gf1_mem_xalloc(alloc, &block) == NULL) return -ENOMEM; #ifdef CONFIG_SND_DEBUG snd_card_ro_proc_new(gus->card, "gusmem", gus, snd_gf1_mem_info_read); --

2 1

Re: [PATCH] ALSA: jack: Check the return value of kstrdup()
by Xiaoke Wang 13 Dec '21

13 Dec '21

kstrdup() can return NULL, it is better to check the return value of it. Fix: free jack and use my full name. Signed-off-by: Xiaoke Wang <xkernel.wang(a)foxmail.com> --- sound/core/jack.c | 4 ++ 1 file changed, 4 insertions(+) diff --git a/sound/core/jack.c b/sound/core/jack.c index 32350c6..5fd9954 100644 --- a/sound/core/jack.c +++ b/sound/core/jack.c @@ -509,6 +509,10 @@ int snd_jack_new(struct snd_card *card, const char *id, int type, return -ENOMEM; jack->id = kstrdup(id, GFP_KERNEL); + if (jack->id == NULL) { + kfree(jack); + return -ENOMEM; + } /* don't creat input device for phantom jack */ if (!phantom_jack) { --

2 1

[PATCH] ALSA: drivers: opl3: Fix incorrect use of vp->state
by Colin Ian King 13 Dec '21

13 Dec '21

Static analysis with scan-build has found an assignment to vp2 that is never used. It seems that the check on vp->state > 0 should be actually on vp2->state instead. Fix this. This dates back to 2002, I found the offending commit from the git history git://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git, commit 91e39521bbf6 ("[PATCH] ALSA patch for 2.5.4") Signed-off-by: Colin Ian King <colin.i.king(a)gmail.com> --- sound/drivers/opl3/opl3_midi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/drivers/opl3/opl3_midi.c b/sound/drivers/opl3/opl3_midi.c index e1b69c65c3c8..e2b7be67f0e3 100644 --- a/sound/drivers/opl3/opl3_midi.c +++ b/sound/drivers/opl3/opl3_midi.c @@ -397,7 +397,7 @@ void snd_opl3_note_on(void *p, int note, int vel, struct snd_midi_channel *chan) } if (instr_4op) { vp2 = &opl3->voices[voice + 3]; - if (vp->state > 0) { + if (vp2->state > 0) { opl3_reg = reg_side | (OPL3_REG_KEYON_BLOCK + voice_offset + 3); reg_val = vp->keyon_reg & ~OPL3_KEYON_BIT; -- 2.33.1

2 1

[PATCH] ALSA: uapi: use C90 comment style instead of C99 style
by Takashi Sakamoto 13 Dec '21

13 Dec '21

UAPI headers are built with compiler option for C90, thus double-slashes comment introduced in C99 is not preferable. Fixes: fb6723daf890 ("ALSA: pcm: comment about relation between msbits hw parameter and [S|U]32 formats") Signed-off-by: Takashi Sakamoto <o-takashi(a)sakamocchi.jp> --- include/uapi/sound/asound.h | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/include/uapi/sound/asound.h b/include/uapi/sound/asound.h index cf1d20e34167..1834f58b8ede 100644 --- a/include/uapi/sound/asound.h +++ b/include/uapi/sound/asound.h @@ -202,9 +202,11 @@ typedef int __bitwise snd_pcm_format_t; #define SNDRV_PCM_FORMAT_S24_BE ((__force snd_pcm_format_t) 7) /* low three bytes */ #define SNDRV_PCM_FORMAT_U24_LE ((__force snd_pcm_format_t) 8) /* low three bytes */ #define SNDRV_PCM_FORMAT_U24_BE ((__force snd_pcm_format_t) 9) /* low three bytes */ -// For S32/U32 formats, 'msbits' hardware parameter is often used to deliver information about the -// available bit count in most significant bit. It's for the case of so-called 'left-justified' or -// `right-padding` sample which has less width than 32 bit. +/* + * For S32/U32 formats, 'msbits' hardware parameter is often used to deliver information about the + * available bit count in most significant bit. It's for the case of so-called 'left-justified' or + * `right-padding` sample which has less width than 32 bit. + */ #define SNDRV_PCM_FORMAT_S32_LE ((__force snd_pcm_format_t) 10) #define SNDRV_PCM_FORMAT_S32_BE ((__force snd_pcm_format_t) 11) #define SNDRV_PCM_FORMAT_U32_LE ((__force snd_pcm_format_t) 12) -- 2.32.0

2 1

[PATCH] ALSA: pcm: comment about relation between msbits hw parameter and [S|U32] formats
by Takashi Sakamoto 13 Dec '21

13 Dec '21

Regarding to handling [U|S][32|24] PCM formats, many userspace application developers and driver developers have confusion, since they require them to understand justification or padding. It easily loses consistency and soundness to operate with many type of devices. In this commit, I attempt to solve the situation by adding comment about relation between [S|U]32 formats and 'msbits' hardware parameter. The formats are used for 'left-justified' sample format, and the available bit count in most significant bit is delivered to userspace in msbits hardware parameter (struct snd_pcm_hw_params.msbits), which is decided by msbits constraint added by pcm drivers (snd_pcm_hw_constraint_msbits()). In driver side, the msbits constraint includes two elements; the physical width of format and the available width of the format in most significant bit. The former is used to match SAMPLE_BITS of format. (For my convenience, I ignore wildcard in the usage of the constraint.) As a result of interaction between ALSA pcm core and ALSA pcm application, when the format in which SAMPLE_BITS equals to physical width of the msbits constaint, the msbits parameter is set by referring to the available width of the constraint. When the msbits parameter is not changed in the above process, ALSA pcm core set it alternatively with SAMPLE_BIT of chosen format. In userspace application side, the msbits is only available after calling ioctl(2) with SNDRV_PCM_IOCTL_HW_PARAMS request. Even if the hardware parameter structure includes somewhat value of SAMPLE_BITS interval parameter as width of format, all of the width is not always available since msbits can be less than the width. I note that [S|U24] formats are used for 'right-justified' 24 bit sample formats within 32 bit frame. The first byte in most significant bit should be invalidated. Although the msbits exposed to userspace should be zero as invalid value, actually it is 32 from physical width of format. Signed-off-by: Takashi Sakamoto <o-takashi(a)sakamocchi.jp> --- include/sound/pcm.h | 3 +++ include/uapi/sound/asound.h | 3 +++ 2 files changed, 6 insertions(+) diff --git a/include/sound/pcm.h b/include/sound/pcm.h index 2e1200d17d0c..b84ff6e4ee53 100644 --- a/include/sound/pcm.h +++ b/include/sound/pcm.h @@ -147,6 +147,9 @@ struct snd_pcm_ops { #define SNDRV_PCM_FMTBIT_S24_BE _SNDRV_PCM_FMTBIT(S24_BE) #define SNDRV_PCM_FMTBIT_U24_LE _SNDRV_PCM_FMTBIT(U24_LE) #define SNDRV_PCM_FMTBIT_U24_BE _SNDRV_PCM_FMTBIT(U24_BE) +// For S32/U32 formats, 'msbits' hardware parameter is often used to deliver information about the +// available bit count in most significant bit. It's for the case of so-called 'left-justified' or +// `right-padding` sample which has less width than 32 bit. #define SNDRV_PCM_FMTBIT_S32_LE _SNDRV_PCM_FMTBIT(S32_LE) #define SNDRV_PCM_FMTBIT_S32_BE _SNDRV_PCM_FMTBIT(S32_BE) #define SNDRV_PCM_FMTBIT_U32_LE _SNDRV_PCM_FMTBIT(U32_LE) diff --git a/include/uapi/sound/asound.h b/include/uapi/sound/asound.h index d17c061950df..b43ea0067e1f 100644 --- a/include/uapi/sound/asound.h +++ b/include/uapi/sound/asound.h @@ -202,6 +202,9 @@ typedef int __bitwise snd_pcm_format_t; #define SNDRV_PCM_FORMAT_S24_BE ((__force snd_pcm_format_t) 7) /* low three bytes */ #define SNDRV_PCM_FORMAT_U24_LE ((__force snd_pcm_format_t) 8) /* low three bytes */ #define SNDRV_PCM_FORMAT_U24_BE ((__force snd_pcm_format_t) 9) /* low three bytes */ +// For S32/U32 formats, 'msbits' hardware parameter is often used to deliver information about the +// available bit count in most significant bit. It's for the case of so-called 'left-justified' or +// `right-padding` sample which has less width than 32 bit. #define SNDRV_PCM_FORMAT_S32_LE ((__force snd_pcm_format_t) 10) #define SNDRV_PCM_FORMAT_S32_BE ((__force snd_pcm_format_t) 11) #define SNDRV_PCM_FORMAT_U32_LE ((__force snd_pcm_format_t) 12) -- 2.27.0

2 5

No audio from Starship/Matisse HD Audio Controller
by GitHub issues - edited 13 Dec '21

13 Dec '21

alsa-project/alsa-ucm-conf issue #129 was edited from SirFooOfBar: I have an AMD ryzen 3900X paired with a Gigabyte AX370-Gaming 5 motherboard. I use arch linux. Upon upgrading the package alsa-ucm-conf from version 1.2.6.1-1 to 1.2.6.2-2 and rebooting, my onboard audio stops working completely. Downgrading restores functionality after another reboot. In pavucontrol, for 1.2.6.1-1 there are many available profiles, but for 1.2.6.2-2 there is only one, which says unavailable. I also use pipewire rather than pulseaudio if that makes a difference Issue URL : https://github.com/alsa-project/alsa-ucm-conf/issues/129 Repository URL: https://github.com/alsa-project/alsa-ucm-conf

1 0