July 2017 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] Slow/piecewise send of sysex?
by Ville Oikarinen 12 Jul '17

12 Jul '17

Hello. I just created this question as a ticket https://github.com/opensrc/alsa/issues/24 but was instructed to ask here. So: I'm using amidi --send-hex to send data to a Commodore 64 via a CLab MIDI interface. The problem is that even with 100 bytes the probability of a buffer overrun is about 30%. I didn't find a way to throttle the send speed. I also tried sending the data in smaller chunks, but I cannot get amidi to send anything after the first chunk. I suspect it refuses to send anything if the data doesn't start with F0 i.e. sysex dump start. If I am right about that assumption, as a workaround I could encapsulate each chunk as a full sysex dump, but that requires extra logic on the receiving side and it also wastes bandwidth, so I don't really wish to do that. Primarily I would prefer a way to throttle the speed, but if I cannot have that, I would like to be able to send the data a few bytes at a time in a simple shell loop. - Ville Oikarinen

2 2

[alsa-devel] [PATCH v2] ASoC: samsung: i2s: Support more resolution rates
by Jaechul Lee 12 Jul '17

12 Jul '17

This driver can support more frequencies over 96KHz. There are no reasons to limit the frequency range below 96KHz. If codecs/amps or something else can't support higher resolution rates, the constraints would be set rates properly because each drivers have its own limits. I added the 'pcm_rates' field to the dai_data to be set rates by the compatibilities. As a result, rates will be set each devices respectively. For example of exynos5433, rates will be set from 8KHz to 192KHz. Signed-off-by: Jaechul Lee <jcsing.lee(a)samsung.com> --- v2: - changed the name of variable to pcm_rates. - removed duplicated code. - modified commit message. --- sound/soc/samsung/i2s.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/sound/soc/samsung/i2s.c b/sound/soc/samsung/i2s.c index af3ba4d4ccc5..c9f87f7bae99 100644 --- a/sound/soc/samsung/i2s.c +++ b/sound/soc/samsung/i2s.c @@ -50,6 +50,7 @@ struct samsung_i2s_variant_regs { struct samsung_i2s_dai_data { u32 quirks; + unsigned int pcm_rates; const struct samsung_i2s_variant_regs *i2s_variant_regs; }; @@ -1076,13 +1077,13 @@ static const struct snd_soc_component_driver samsung_i2s_component = { .name = "samsung-i2s", }; -#define SAMSUNG_I2S_RATES SNDRV_PCM_RATE_8000_96000 - #define SAMSUNG_I2S_FMTS (SNDRV_PCM_FMTBIT_S8 | \ SNDRV_PCM_FMTBIT_S16_LE | \ SNDRV_PCM_FMTBIT_S24_LE) -static struct i2s_dai *i2s_alloc_dai(struct platform_device *pdev, bool sec) +static struct i2s_dai *i2s_alloc_dai(struct platform_device *pdev, + const struct samsung_i2s_dai_data *i2s_dai_data, + bool sec) { struct i2s_dai *i2s; @@ -1101,13 +1102,13 @@ static struct i2s_dai *i2s_alloc_dai(struct platform_device *pdev, bool sec) i2s->i2s_dai_drv.resume = i2s_resume; i2s->i2s_dai_drv.playback.channels_min = 1; i2s->i2s_dai_drv.playback.channels_max = 2; - i2s->i2s_dai_drv.playback.rates = SAMSUNG_I2S_RATES; + i2s->i2s_dai_drv.playback.rates = i2s_dai_data->pcm_rates; i2s->i2s_dai_drv.playback.formats = SAMSUNG_I2S_FMTS; if (!sec) { i2s->i2s_dai_drv.capture.channels_min = 1; i2s->i2s_dai_drv.capture.channels_max = 2; - i2s->i2s_dai_drv.capture.rates = SAMSUNG_I2S_RATES; + i2s->i2s_dai_drv.capture.rates = i2s_dai_data->pcm_rates; i2s->i2s_dai_drv.capture.formats = SAMSUNG_I2S_FMTS; } return i2s; @@ -1242,7 +1243,7 @@ static int samsung_i2s_probe(struct platform_device *pdev) i2s_dai_data = (struct samsung_i2s_dai_data *) platform_get_device_id(pdev)->driver_data; - pri_dai = i2s_alloc_dai(pdev, false); + pri_dai = i2s_alloc_dai(pdev, i2s_dai_data, false); if (!pri_dai) { dev_err(&pdev->dev, "Unable to alloc I2S_pri\n"); return -ENOMEM; @@ -1316,7 +1317,7 @@ static int samsung_i2s_probe(struct platform_device *pdev) goto err_disable_clk; if (quirks & QUIRK_SEC_DAI) { - sec_dai = i2s_alloc_dai(pdev, true); + sec_dai = i2s_alloc_dai(pdev, i2s_dai_data, true); if (!sec_dai) { dev_err(&pdev->dev, "Unable to alloc I2S_sec\n"); ret = -ENOMEM; @@ -1452,29 +1453,34 @@ static const struct samsung_i2s_variant_regs i2sv5_i2s1_regs = { static const struct samsung_i2s_dai_data i2sv3_dai_type = { .quirks = QUIRK_NO_MUXPSR, + .pcm_rates = SNDRV_PCM_RATE_8000_96000, .i2s_variant_regs = &i2sv3_regs, }; static const struct samsung_i2s_dai_data i2sv5_dai_type = { .quirks = QUIRK_PRI_6CHAN | QUIRK_SEC_DAI | QUIRK_NEED_RSTCLR | QUIRK_SUPPORTS_IDMA, + .pcm_rates = SNDRV_PCM_RATE_8000_96000, .i2s_variant_regs = &i2sv3_regs, }; static const struct samsung_i2s_dai_data i2sv6_dai_type = { .quirks = QUIRK_PRI_6CHAN | QUIRK_SEC_DAI | QUIRK_NEED_RSTCLR | QUIRK_SUPPORTS_TDM | QUIRK_SUPPORTS_IDMA, + .pcm_rates = SNDRV_PCM_RATE_8000_96000, .i2s_variant_regs = &i2sv6_regs, }; static const struct samsung_i2s_dai_data i2sv7_dai_type = { .quirks = QUIRK_PRI_6CHAN | QUIRK_SEC_DAI | QUIRK_NEED_RSTCLR | QUIRK_SUPPORTS_TDM, + .pcm_rates = SNDRV_PCM_RATE_8000_192000, .i2s_variant_regs = &i2sv7_regs, }; static const struct samsung_i2s_dai_data i2sv5_dai_type_i2s1 = { .quirks = QUIRK_PRI_6CHAN | QUIRK_NEED_RSTCLR, + .pcm_rates = SNDRV_PCM_RATE_8000_96000, .i2s_variant_regs = &i2sv5_i2s1_regs, }; -- 2.11.0

5 4

[alsa-devel] [PATCH - alsa-lib 1/1] snd_user_file: avoid use wordexp
by Natanael Copa 12 Jul '17

12 Jul '17

As suggested in POSIX[1], wordexp might execute the shell. If the libc implementation does so, it will break the firefox sandbox which does not allow exec. This happened on Alpine Linux with musl libc[2]. Since we cannot guarantee that the system wordexp implementation does not execute shell, we cannot really use it, and need to implement the ~/ expansion ourselves. Generally, wordexp is a large attack vector and it is better to avoid it since only tilde expansion is needed. [1]: http://pubs.opengroup.org/onlinepubs/9699919799/functions/wordexp.html#tag_… [2]: http://bugs.alpinelinux.org/issues/7454#note-2 Signed-off-by: Natanael Copa <ncopa(a)alpinelinux.org> diff --git a/configure.ac b/configure.ac index 26e5d125..076c6bab 100644 --- a/configure.ac +++ b/configure.ac @@ -304,7 +304,7 @@ fi AC_SUBST(ALSA_DEPLIBS) dnl Check for headers -AC_CHECK_HEADERS([wordexp.h endian.h sys/endian.h sys/shm.h]) +AC_CHECK_HEADERS([endian.h sys/endian.h sys/shm.h]) dnl Check for resmgr support... AC_MSG_CHECKING(for resmgr support) diff --git a/src/userfile.c b/src/userfile.c index 72779da4..0e3f5fae 100644 --- a/src/userfile.c +++ b/src/userfile.c @@ -21,6 +21,11 @@ #include <config.h> #include <string.h> #include <errno.h> +#include <sys/types.h> +#include <unistd.h> +#include <pwd.h> +#include <stdio.h> +#include <stdlib.h> /** * \brief Get the full file name @@ -28,46 +33,58 @@ * \param result The pointer to store the resultant file name * \return 0 if successful, or a negative error code * - * Parses the given file name with POSIX-Shell-like expansion and - * stores the first matchine one. The returned string is strdup'ed. + * Parses the given file name with POSIX-Shell-like expansion for ~/. + * The returned string is strdup'ed. */ -#ifdef HAVE_WORDEXP_H -#include <wordexp.h> #include <assert.h> int snd_user_file(const char *file, char **result) { - wordexp_t we; int err; - + size_t len; + char *buf = NULL; + assert(file && result); - err = wordexp(file, &we, WRDE_NOCMD); - switch (err) { - case WRDE_NOSPACE: - wordfree(&we); - return -ENOMEM; - case 0: - if (we.we_wordc == 1) - break; - wordfree(&we); - /* fall thru */ - default: - return -EINVAL; + *result = NULL; + + /* expand ~/ if needed */ + if (file[0] == '~' && file[1] == '/') { + const char *home = getenv("HOME"); + if (home == NULL) { + struct passwd pwent, *p = NULL; + uid_t id = getuid(); + size_t bufsize = 1024; + + buf = malloc(bufsize); + if (buf == NULL) + goto out; + + while ((err = getpwuid_r(id, &pwent, buf, bufsize, &p)) == ERANGE) { + char *newbuf; + bufsize += 1024; + if (bufsize < 1024) + break; + newbuf = realloc(buf, bufsize); + if (newbuf == NULL) + goto out; + buf = newbuf; + } + home = err ? "" : pwent.pw_dir; + } + len = strlen(home) + strlen(&file[2]) + 2; + *result = malloc(len); + if (*result) + snprintf(*result, len, "%s/%s", home, &file[2]); + } else { + *result = strdup(file); } - *result = strdup(we.we_wordv[0]); - wordfree(&we); + +out: + if (buf) + free(buf); + if (*result == NULL) return -ENOMEM; return 0; } -#else /* !HAVE_WORDEXP_H */ -/* just copy the string - would be nicer to expand by ourselves, though... */ -int snd_user_file(const char *file, char **result) -{ - *result = strdup(file); - if (! *result) - return -ENOMEM; - return 0; -} -#endif /* HAVE_WORDEXP_H */ -- 2.13.2

1 0

[alsa-devel] Clock Drift between hardware sound card and snd_aloop
by Rob Nertney 12 Jul '17

12 Jul '17

Hi all, I have successfully built a sound capture card that runs at a given rate off its own clock. I receive the stream and do some post-processing on it, and I want to present this modified data to the user as a capture pcm. I initially thought that snd_aloop would be the best idea, as I can pipe my post-processed data directly into it. However, the driver is apparently calculating exact timing of what "should" be the rate based off of CPU ticks/jiffies and the requested period/sample rate/etc in the requested output stream. In my actual hardware I see cycle slips. Underruns and overruns based off of minor adjustments to the card's rate. Is snd_aloop the appropriate driver to be using? Or should I create a new card, similar to snd_aloop, which is more specifically interrupted (and associated pointer locations, etc. modified) from the hardware card as well? It seems like I should be able to synchronize snd_aloop's timing with my hardware clock, but I haven't found anywhere in the API to achieve this. How else can I avoid this clock drift? Thanks, Rob

2 1

[alsa-devel] [PATCH] WIP:asoc:qdsp6: fix semicolon.cocci warnings
by Julia Lawall 11 Jul '17

11 Jul '17

Remove unneeded semicolons. Generated by: scripts/coccinelle/misc/semicolon.cocci Signed-off-by: Julia Lawall <julia.lawall(a)lip6.fr> Signed-off-by: Fengguang Wu <fengguang.wu(a)intel.com> --- tree: https://github.com/rrnayak/linux genpd-performance-state head: 3f1a0555478c46892595728fcb4a2c7ae7e1f7da commit: b4f8ada6fb36369cf0e45421dde636d79e845554 [48/114] WIP:asoc:qdsp6: add adm driver :::::: branch date: 2 hours ago :::::: commit date: 8 weeks ago q6adm.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- a/sound/soc/qcom/qdsp6v2/q6adm.c +++ b/sound/soc/qcom/qdsp6v2/q6adm.c @@ -806,7 +806,7 @@ int adm_arrange_mch_map(struct adm_cmd_d default: goto non_mch_path; break; - }; + } if ((open->dev_num_channel > 2) && multi_ch_maps[idx].set_channel_map) memcpy(open->dev_channel_mapping, @@ -1019,10 +1019,10 @@ int adm_matrix_map(int path, struct rout route->hdr.pkt_size = cmd_size; route->hdr.src_svc = 0; route->hdr.src_domain = APR_DOMAIN_APPS; - route->hdr.src_port = 0; /* Ignored */; + route->hdr.src_port = 0; /* Ignored */ route->hdr.dest_svc = APR_SVC_ADM; route->hdr.dest_domain = APR_DOMAIN_ADSP; - route->hdr.dest_port = 0; /* Ignored */; + route->hdr.dest_port = 0; /* Ignored */ route->hdr.token = 0; if (path == ADM_PATH_COMPRESSED_RX) { pr_debug("%s: ADM_CMD_STREAM_DEVICE_MAP_ROUTINGS_V5 0x%x\n",

1 0

[alsa-devel] [PATCH] ASoC: pxa: SND_PXA2XX_SOC should depend on HAS_DMA
by Geert Uytterhoeven 11 Jul '17

11 Jul '17

If NO_DMA=y: ERROR: "bad_dma_ops" [sound/soc/pxa/snd-soc-pxa2xx.ko] undefined! ERROR: "bad_dma_ops" [sound/arm/snd-pxa2xx-lib.ko] undefined! ERROR: "dma_common_mmap" [sound/arm/snd-pxa2xx-lib.ko] undefined! Add a dependency on HAS_DMA to fix this. Fixes: 73d7ee2e831f106c ("ASoC: pxa: add COMPILE_TEST on SND_PXA2XX_SOC") Signed-off-by: Geert Uytterhoeven <geert(a)linux-m68k.org> --- sound/soc/pxa/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/pxa/Kconfig b/sound/soc/pxa/Kconfig index 960744e46edc0549..484ab3c2ad672fc8 100644 --- a/sound/soc/pxa/Kconfig +++ b/sound/soc/pxa/Kconfig @@ -1,6 +1,7 @@ config SND_PXA2XX_SOC tristate "SoC Audio for the Intel PXA2xx chip" depends on ARCH_PXA || COMPILE_TEST + depends on HAS_DMA select SND_PXA2XX_LIB help Say Y or M if you want to add support for codecs attached to -- 2.7.4

3 2

[alsa-devel] Applied "ASoC: hdmi-codec: make const array hdmi_codec_eld_spk_alloc_bits static" to the asoc tree
by Mark Brown 11 Jul '17

11 Jul '17

The patch ASoC: hdmi-codec: make const array hdmi_codec_eld_spk_alloc_bits static has been applied to the asoc tree at git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted. You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed. If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced. Please add any relevant lists and maintainers to the CCs when replying to this mail. Thanks, Mark >From fc2b74460d96ec28110a581d6e58d15c27304d16 Mon Sep 17 00:00:00 2001 From: Colin Ian King <colin.king(a)canonical.com> Date: Tue, 11 Jul 2017 15:08:55 +0100 Subject: [PATCH] ASoC: hdmi-codec: make const array hdmi_codec_eld_spk_alloc_bits static Don't populate array hdmi_codec_eld_spk_alloc_bits on the stack but make it static. Makes the object code smaller by over 260 bytes: Before: text data bss dec hex filename 10882 3384 64 14330 37fa sound/soc/codecs/hdmi-codec.o After: text data bss dec hex filename 10557 3440 64 14061 36ed sound/soc/codecs/hdmi-codec.o Signed-off-by: Colin Ian King <colin.king(a)canonical.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> --- sound/soc/codecs/hdmi-codec.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/hdmi-codec.c b/sound/soc/codecs/hdmi-codec.c index 7686a80861f1..509ab513b4b2 100644 --- a/sound/soc/codecs/hdmi-codec.c +++ b/sound/soc/codecs/hdmi-codec.c @@ -326,7 +326,7 @@ static int hdmi_eld_ctl_get(struct snd_kcontrol *kcontrol, static unsigned long hdmi_codec_spk_mask_from_alloc(int spk_alloc) { int i; - const unsigned long hdmi_codec_eld_spk_alloc_bits[] = { + static const unsigned long hdmi_codec_eld_spk_alloc_bits[] = { [0] = FL | FR, [1] = LFE, [2] = FC, [3] = RL | RR, [4] = RC, [5] = FLC | FRC, [6] = RLC | RRC, }; -- 2.13.2

1 0

[alsa-devel] [PATCH] ASoC: hdmi-codec: make const array hdmi_codec_eld_spk_alloc_bits static
by Colin King 11 Jul '17

11 Jul '17

From: Colin Ian King <colin.king(a)canonical.com> Don't populate array hdmi_codec_eld_spk_alloc_bits on the stack but make it static. Makes the object code smaller by over 260 bytes: Before: text data bss dec hex filename 10882 3384 64 14330 37fa sound/soc/codecs/hdmi-codec.o After: text data bss dec hex filename 10557 3440 64 14061 36ed sound/soc/codecs/hdmi-codec.o Signed-off-by: Colin Ian King <colin.king(a)canonical.com> --- sound/soc/codecs/hdmi-codec.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/hdmi-codec.c b/sound/soc/codecs/hdmi-codec.c index 7686a80861f1..509ab513b4b2 100644 --- a/sound/soc/codecs/hdmi-codec.c +++ b/sound/soc/codecs/hdmi-codec.c @@ -326,7 +326,7 @@ static int hdmi_eld_ctl_get(struct snd_kcontrol *kcontrol, static unsigned long hdmi_codec_spk_mask_from_alloc(int spk_alloc) { int i; - const unsigned long hdmi_codec_eld_spk_alloc_bits[] = { + static const unsigned long hdmi_codec_eld_spk_alloc_bits[] = { [0] = FL | FR, [1] = LFE, [2] = FC, [3] = RL | RR, [4] = RC, [5] = FLC | FRC, [6] = RLC | RRC, }; -- 2.11.0

1 0

[alsa-devel] New codec device ID for ALC1220
by Kailang 11 Jul '17

11 Jul '17

Hi Takashi, ALC1220 has another device ID. Attach patch was add 0x1168 device ID support. Many Thanks. BR, Kailang

2 1

[alsa-devel] [PATCH][-next] ASoC: rt5663: add in missing loop counter to avoid infinite loop
by Colin King 10 Jul '17

10 Jul '17

From: Colin Ian King <colin.king(a)canonical.com> The loop counter 'count' is missing an increment, so we could end up with an infinite loop. Add in a pre-increment to count to fix this. Detected by CoverityScan, CID#1450120 ("Logically Dead Code") Fixes: 1a425dbf1a10 ("ASoC: rt5663: Update the calibration funciton") Signed-off-by: Colin Ian King <colin.king(a)canonical.com> --- sound/soc/codecs/rt5663.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/rt5663.c b/sound/soc/codecs/rt5663.c index 82edbcf39caa..5b3c50c641d8 100644 --- a/sound/soc/codecs/rt5663.c +++ b/sound/soc/codecs/rt5663.c @@ -3044,7 +3044,7 @@ static void rt5663_calibrate(struct rt5663_priv *rt5663) else break; - if (count > 200) + if (++count > 200) break; } -- 2.11.0

2 1