January 2018 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] Where do amidi MIDI port names come from ?
by Happy 22 Jan '18

22 Jan '18

Does someone know where (USB) MIDI port names in Linux come from ? For example a MIDISport 4x4 shows "Midiman" at lsusb (the same for each one if there is more than one device. This comes from the idVendor, idProduct in the device descriptor) But "MIDISPORT 4x4 Anniv MIDI 1 to 4" by amidi -l for the device. Thus wonder where amidi (ALSA) gets the product/portname from. There are other examples where the SUB device ID is not known by Linux but the port name does have the vendor/product name Microsoft argues that some MIDI devices do not report a name But since Linux does not have a driver for each device I highly doubt so. (have not come across one) Microsoft chose to use the name from the device driver in their UWP MIDI implementation which has apparantly been not a good idea It has resulted in multiple duplicate device/port names for different devices.

2 1

[alsa-devel] Maximum 50 dB gain in ALSA softvol plugin
by Ricard Wanderlof 22 Jan '18

22 Jan '18

Hi, The ALSA softvol plugin has a maximum gain of 50 dB (MAX_DB_UPPER_LIMIT is fixed to 50 in pcm/pcm_softvol.c). Is there some technical reason for this limit? These days when >16 bit audio is becoming common, it's not unreasonable to have large gain figures, and that is in fact a usecase I have right now. I'm looking through the CONVERT_AREA and MULTI_DIV_foo macros to see if there's something there that degenerates when the gain factor becomes too large, but I'd thought I'd ask too rather than just do reverse engineering. /Ricard -- Ricard Wolf Wanderlöf ricardw(at)axis.com Axis Communications AB, Lund, Sweden www.axis.com Phone +46 46 272 2016 Fax +46 46 13 61 30

3 10

[alsa-devel] [PATCH] ASoC: sgtl5000: Clarify a bit about the ER1 meaning
by Fabio Estevam 22 Jan '18

22 Jan '18

The "check ER1" message is not very clear about its meaning. Improve it a bit by referring to it as "ER1 erratum" so that it becomes clearer that ER1 references to a SGTL5000 erratum. Signed-off-by: Fabio Estevam <fabio.estevam(a)nxp.com> --- sound/soc/codecs/sgtl5000.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/sgtl5000.c b/sound/soc/codecs/sgtl5000.c index 633cdcf..e1ab553 100644 --- a/sound/soc/codecs/sgtl5000.c +++ b/sound/soc/codecs/sgtl5000.c @@ -1392,7 +1392,7 @@ static int sgtl5000_i2c_probe(struct i2c_client *client, ana_pwr |= SGTL5000_LINEREG_D_POWERUP; dev_info(&client->dev, - "Using internal LDO instead of VDDD: check ER1\n"); + "Using internal LDO instead of VDDD: check ER1 erratum\n"); } else { /* using external LDO for VDDD * Clear startup powerup and simple powerup -- 2.7.4

2 1

[alsa-devel] [PATCH] ASoC: mediatek: fix double free in mt2701_afe_pcm_dev_probe()
by Ryder Lee 22 Jan '18

22 Jan '18

The commit dfa3cbb8(ASoC: mediatek: modify MT2701 AFE driver to adapt mfd device) leads to the following static checker warning: sound/soc/mediatek/mt2701/mt2701-afe-pcm.c:1535 mt2701_afe_pcm_dev_probe() error: double free of 'component' This patch fixes that and adds a helper mt2701_afe_add_component() to setup component. Reported-by: Dan Carpenter <dan.carpenter(a)oracle.com> Signed-off-by: Ryder Lee <ryder.lee(a)mediatek.com> --- sound/soc/mediatek/mt2701/mt2701-afe-pcm.c | 32 +++++++++++++++++------------- 1 file changed, 18 insertions(+), 14 deletions(-) diff --git a/sound/soc/mediatek/mt2701/mt2701-afe-pcm.c b/sound/soc/mediatek/mt2701/mt2701-afe-pcm.c index f0cd08f..ff15537 100644 --- a/sound/soc/mediatek/mt2701/mt2701-afe-pcm.c +++ b/sound/soc/mediatek/mt2701/mt2701-afe-pcm.c @@ -1405,9 +1405,24 @@ static int mt2701_afe_runtime_resume(struct device *dev) return mt2701_afe_enable_clock(afe); } -static int mt2701_afe_pcm_dev_probe(struct platform_device *pdev) +static int mt2701_afe_add_component(struct mtk_base_afe *afe) { struct snd_soc_component *component; + + component = kzalloc(sizeof(*component), GFP_KERNEL); + if (!component) + return -ENOMEM; + + component->regmap = afe->regmap; + + return snd_soc_add_component(afe->dev, component, + &mt2701_afe_pcm_dai_component, + mt2701_afe_pcm_dais, + ARRAY_SIZE(mt2701_afe_pcm_dais)); +} + +static int mt2701_afe_pcm_dev_probe(struct platform_device *pdev) +{ struct mtk_base_afe *afe; struct mt2701_afe_private *afe_priv; struct device *dev; @@ -1477,12 +1492,6 @@ static int mt2701_afe_pcm_dev_probe(struct platform_device *pdev) = &mt2701_i2s_data[i][I2S_IN]; } - component = kzalloc(sizeof(*component), GFP_KERNEL); - if (!component) - return -ENOMEM; - - component->regmap = afe->regmap; - afe->mtk_afe_hardware = &mt2701_afe_hardware; afe->memif_fs = mt2701_memif_fs; afe->irq_fs = mt2701_irq_fs; @@ -1495,7 +1504,7 @@ static int mt2701_afe_pcm_dev_probe(struct platform_device *pdev) ret = mt2701_init_clock(afe); if (ret) { dev_err(dev, "init clock error\n"); - goto err_init_clock; + return ret; } platform_set_drvdata(pdev, afe); @@ -1514,10 +1523,7 @@ static int mt2701_afe_pcm_dev_probe(struct platform_device *pdev) goto err_platform; } - ret = snd_soc_add_component(dev, component, - &mt2701_afe_pcm_dai_component, - mt2701_afe_pcm_dais, - ARRAY_SIZE(mt2701_afe_pcm_dais)); + ret = mt2701_afe_add_component(afe); if (ret) { dev_warn(dev, "err_dai_component\n"); goto err_dai_component; @@ -1531,8 +1537,6 @@ static int mt2701_afe_pcm_dev_probe(struct platform_device *pdev) pm_runtime_put_sync(dev); err_pm_disable: pm_runtime_disable(dev); -err_init_clock: - kfree(component); return ret; } -- 1.9.1

2 1

[alsa-devel] [PATCH] ASoC: codecs: dmic: Fix check of return value from read of 'num-channels'
by Matthias Kaehlcke 22 Jan '18

22 Jan '18

Commit 7fb59e940f62 ("ASoC: codecs: dmic: Make number of channels configurable") introduces an optional property to the device tree to specify the number of DMIC channels. dmic_codec_probe() uses of_property_read_u32() to read the DT value, and expects a return value of -ENOENT when the property does not exist. This expectation is incorrect, the actual value returned in this case is -EINVAL (see of_find_property_value_of_size(), which is called under the hood). Check for -EINVAL instead. Fixes: 7fb59e940f62 ("ASoC: codecs: dmic: Make number of channels configurable") Signed-off-by: Matthias Kaehlcke <mka(a)chromium.org> --- sound/soc/codecs/dmic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/dmic.c b/sound/soc/codecs/dmic.c index c88f974ebe3e..cf83c423394d 100644 --- a/sound/soc/codecs/dmic.c +++ b/sound/soc/codecs/dmic.c @@ -113,7 +113,7 @@ static int dmic_dev_probe(struct platform_device *pdev) if (pdev->dev.of_node) { err = of_property_read_u32(pdev->dev.of_node, "num-channels", &chans); - if (err && (err != -ENOENT)) + if (err && (err != -EINVAL)) return err; if (!err) { -- 2.16.0.rc1.238.g530d649a79-goog

2 2

[alsa-devel] [PATCH] ASoC: max98373 Changed SPDX header in C++ comments style
by Ryan Lee 22 Jan '18

22 Jan '18

Signed-off-by: Ryan Lee <ryans.lee(a)maximintegrated.com> --- Chagelog: Changed SPDX header in C comment-style to C++ comment style sound/soc/codecs/max98373.c | 4 ++-- sound/soc/codecs/max98373.h | 5 +++-- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/max98373.c b/sound/soc/codecs/max98373.c index 31b0864..562e887 100644 --- a/sound/soc/codecs/max98373.c +++ b/sound/soc/codecs/max98373.c @@ -1,5 +1,5 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* Copyright (c) 2017, Maxim Integrated */ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (c) 2017, Maxim Integrated #include <linux/acpi.h> #include <linux/i2c.h> diff --git a/sound/soc/codecs/max98373.h b/sound/soc/codecs/max98373.h index d0b359d..f6a37aa 100644 --- a/sound/soc/codecs/max98373.h +++ b/sound/soc/codecs/max98373.h @@ -1,5 +1,6 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* Copyright (c) 2017, Maxim Integrated */ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (c) 2017, Maxim Integrated + #ifndef _MAX98373_H #define _MAX98373_H -- 2.7.4

2 1

[alsa-devel] [PATCH 0/5] ASoC: (minor) atmel cleanup
by Ladislav Michl 22 Jan '18

22 Jan '18

Hi there! this patchset is a side product of reading ALSA code while trying to support new hardware. Nothing special here, just an attempt to make those three drivers to look more similar. Only patch 4 fixes actual problem, but: 1) this particular issue is present almost everywhere (as far as I looked) where of_node_put is used in ALSA. Perhaps coccinelle people could help? 2) seems noone bothered as OF_DYNAMIC is mostly disabled Compile tested only. Ladislav Michl (5): ASoC: sam9g20_wm8731: use dev_*() logging functions ASoC: sam9x5_wm8731: Drop 'ASoC' prefix from error messages ASoC: sam9x5_wm8731: Return -ENODEV when probe does not find OF node ASoC: sam9x5_wm8731: Fix OF node reference counting ASoC: atmel_wm8904: Return -ENODEV when probe does not find OF node sound/soc/atmel/atmel_wm8904.c | 6 ++---- sound/soc/atmel/sam9g20_wm8731.c | 17 +++++++-------- sound/soc/atmel/sam9x5_wm8731.c | 45 +++++++++++++++++----------------------- 3 files changed, 29 insertions(+), 39 deletions(-) -- 2.15.1

3 12

[alsa-devel] [bug report] ASoC: mediatek: modify MT2701 AFE driver to adapt mfd device
by Dan Carpenter 22 Jan '18

22 Jan '18

Hello Ryder Lee, The patch dfa3cbb83e09: "ASoC: mediatek: modify MT2701 AFE driver to adapt mfd device" from Jan 4, 2018, leads to the following static checker warning: sound/soc/mediatek/mt2701/mt2701-afe-pcm.c:1535 mt2701_afe_pcm_dev_probe() error: double free of 'component' sound/soc/mediatek/mt2701/mt2701-afe-pcm.c 1509 pm_runtime_get_sync(dev); 1510 1511 ret = snd_soc_register_platform(dev, &mtk_afe_pcm_platform); 1512 if (ret) { 1513 dev_warn(dev, "err_platform\n"); 1514 goto err_platform; 1515 } 1516 1517 ret = snd_soc_add_component(dev, component, ^^^^^^^^^ snd_soc_add_component() frees "component" on error. This seems like a layering violation to me, and it naturally leads to bugs like this. 1518 &mt2701_afe_pcm_dai_component, 1519 mt2701_afe_pcm_dais, 1520 ARRAY_SIZE(mt2701_afe_pcm_dais)); 1521 if (ret) { 1522 dev_warn(dev, "err_dai_component\n"); 1523 goto err_dai_component; 1524 } 1525 1526 return 0; 1527 1528 err_dai_component: 1529 snd_soc_unregister_platform(dev); 1530 err_platform: 1531 pm_runtime_put_sync(dev); 1532 err_pm_disable: 1533 pm_runtime_disable(dev); 1534 err_init_clock: 1535 kfree(component); ^^^^^^^^^ Double free 1536 1537 return ret; 1538 } regards, dan carpenter

2 1

[alsa-devel] [PATCH 3/5] ALSA: emu10k1: add optional debug printouts with DMA addresses
by Maciej S. Szmigiero 21 Jan '18

21 Jan '18

When we get a IOMMU page fault for a emu10k1 device it is very hard to discover which of chip many DMA allocations triggered it (since on a IOMMU system the DMA address space is often very different from the CPU one). Let's add optional debug printouts providing this information. These debug printouts are only enabled on an explicit request via the kernel dynamic debug mechanism. Signed-off-by: Maciej S. Szmigiero <mail(a)maciej.szmigiero.name> --- sound/pci/emu10k1/emu10k1_main.c | 8 ++++++++ sound/pci/emu10k1/memory.c | 11 +++++++++-- 2 files changed, 17 insertions(+), 2 deletions(-) diff --git a/sound/pci/emu10k1/emu10k1_main.c b/sound/pci/emu10k1/emu10k1_main.c index 11e868f569d6..8decd2a7a404 100644 --- a/sound/pci/emu10k1/emu10k1_main.c +++ b/sound/pci/emu10k1/emu10k1_main.c @@ -1898,6 +1898,9 @@ int snd_emu10k1_create(struct snd_card *card, err = -ENOMEM; goto error; } + dev_dbg(card->dev, "page table address range is %.8lx:%.8lx\n", + (unsigned long)emu->ptb_pages.addr, + (unsigned long)(emu->ptb_pages.addr + emu->ptb_pages.bytes)); emu->page_ptr_table = vmalloc(emu->max_cache_pages * sizeof(void *)); emu->page_addr_table = vmalloc(emu->max_cache_pages * @@ -1912,6 +1915,11 @@ int snd_emu10k1_create(struct snd_card *card, err = -ENOMEM; goto error; } + dev_dbg(card->dev, "silent page range is %.8lx:%.8lx\n", + (unsigned long)emu->silent_page.addr, + (unsigned long)(emu->silent_page.addr + + emu->silent_page.bytes)); + emu->memhdr = snd_util_memhdr_new(emu->max_cache_pages * PAGE_SIZE); if (emu->memhdr == NULL) { err = -ENOMEM; diff --git a/sound/pci/emu10k1/memory.c b/sound/pci/emu10k1/memory.c index eaa61024ac7f..39afe199e1c7 100644 --- a/sound/pci/emu10k1/memory.c +++ b/sound/pci/emu10k1/memory.c @@ -35,6 +35,8 @@ */ #define __set_ptb_entry(emu,page,addr) \ (((u32 *)(emu)->ptb_pages.area)[page] = cpu_to_le32(((addr) << (emu->address_mode)) | (page))) +#define __get_ptb_entry(emu, page) \ + (le32_to_cpu(((u32 *)(emu)->ptb_pages.area)[page])) #define UNIT_PAGES (PAGE_SIZE / EMUPAGESIZE) #define MAX_ALIGN_PAGES0 (MAXPAGES0 / UNIT_PAGES) @@ -44,7 +46,7 @@ /* get offset address from aligned page */ #define aligned_page_offset(page) ((page) << PAGE_SHIFT) -#if PAGE_SIZE == 4096 +#if PAGE_SIZE == 4096 && !IS_ENABLED(CONFIG_DYNAMIC_DEBUG) /* page size == EMUPAGESIZE */ /* fill PTB entrie(s) corresponding to page with addr */ #define set_ptb_entry(emu,page,addr) __set_ptb_entry(emu,page,addr) @@ -58,6 +60,8 @@ static inline void set_ptb_entry(struct snd_emu10k1 *emu, int page, dma_addr_t a page *= UNIT_PAGES; for (i = 0; i < UNIT_PAGES; i++, page++) { __set_ptb_entry(emu, page, addr); + dev_dbg(emu->card->dev, "mapped page %d to entry %.8x\n", page, + (unsigned int)__get_ptb_entry(emu, page)); addr += EMUPAGESIZE; } } @@ -65,9 +69,12 @@ static inline void set_silent_ptb(struct snd_emu10k1 *emu, int page) { int i; page *= UNIT_PAGES; - for (i = 0; i < UNIT_PAGES; i++, page++) + for (i = 0; i < UNIT_PAGES; i++, page++) { /* do not increment ptr */ __set_ptb_entry(emu, page, emu->silent_page.addr); + dev_dbg(emu->card->dev, "mapped silent page %d to entry %.8x\n", + page, (unsigned int)__get_ptb_entry(emu, page)); + } } #endif /* PAGE_SIZE */

2 1

[alsa-devel] [PATCH 5/5] ALSA: emu10k1: add a IOMMU workaround
by Maciej S. Szmigiero 21 Jan '18

21 Jan '18

The Audigy 2 CA0102 chip (but most likely others from the emu10k1 family, too) has a problem that from time to time it likes to do few DMA reads a bit beyond its normal allocation and gets very confused if these reads get blocked by a IOMMU. For the first (reserved) page this happens multiple times at every playback, for various synth pages it happens randomly, rarely for the page table memory itself. All these reads seem to follow a similar pattern, observed read offsets beyond the allocation end were 0x00, 0x40, 0x80 and 0xc0 (PCI cache line multiples), so it looks like the device tries to accesses up to 256 extra bytes. As a workaround let's widen these DMA allocations by an extra page if we detect that the device is behind a non-passthrough IOMMU (the DMA memory should be relatively plenty on IOMMU systems). Signed-off-by: Maciej S. Szmigiero <mail(a)maciej.szmigiero.name> --- include/sound/emu10k1.h | 1 + sound/pci/emu10k1/emu10k1_main.c | 49 +++++++++++++++++++++++++++++++++++++--- sound/pci/emu10k1/memory.c | 16 ++++++++++--- 3 files changed, 60 insertions(+), 6 deletions(-) diff --git a/include/sound/emu10k1.h b/include/sound/emu10k1.h index db32b7de52e0..ba27abf65408 100644 --- a/include/sound/emu10k1.h +++ b/include/sound/emu10k1.h @@ -1710,6 +1710,7 @@ struct snd_emu10k1 { unsigned int ecard_ctrl; /* ecard control bits */ unsigned int address_mode; /* address mode */ unsigned long dma_mask; /* PCI DMA mask */ + bool iommu_workaround; /* IOMMU workaround needed */ unsigned int delay_pcm_irq; /* in samples */ int max_cache_pages; /* max memory size / PAGE_SIZE */ struct snd_dma_buffer silent_page; /* silent page */ diff --git a/sound/pci/emu10k1/emu10k1_main.c b/sound/pci/emu10k1/emu10k1_main.c index 8decd2a7a404..3d1b57546379 100644 --- a/sound/pci/emu10k1/emu10k1_main.c +++ b/sound/pci/emu10k1/emu10k1_main.c @@ -36,6 +36,7 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/interrupt.h> +#include <linux/iommu.h> #include <linux/pci.h> #include <linux/slab.h> #include <linux/vmalloc.h> @@ -1758,6 +1759,37 @@ static struct snd_emu_chip_details emu_chip_details[] = { { } /* terminator */ }; +/* + * The chip (at least the Audigy 2 CA0102 chip, but most likely others, too) + * has a problem that from time to time it likes to do few DMA reads a bit + * beyond its normal allocation and gets very confused if these reads get + * blocked by a IOMMU. + * + * This behaviour has been observed for the first (reserved) page + * (for which it happens multiple times at every playback), often for various + * synth pages and sometimes for the page table memory itself. + * + * As a workaround let's widen these DMA allocations by an extra page if we + * detect that the device is behind a non-passthrough IOMMU. + */ +static void snd_emu10k1_detect_iommu(struct snd_emu10k1 *emu) +{ + struct iommu_domain *domain; + + emu->iommu_workaround = false; + + if (!iommu_present(emu->card->dev->bus)) + return; + + domain = iommu_get_domain_for_dev(emu->card->dev); + if (domain && domain->type == IOMMU_DOMAIN_IDENTITY) + return; + + dev_notice(emu->card->dev, + "non-passthrough IOMMU detected, widening DMA allocations"); + emu->iommu_workaround = true; +} + int snd_emu10k1_create(struct snd_card *card, struct pci_dev *pci, unsigned short extin_mask, @@ -1770,6 +1802,7 @@ int snd_emu10k1_create(struct snd_card *card, struct snd_emu10k1 *emu; int idx, err; int is_audigy; + size_t page_table_size, silent_page_size; unsigned int silent_page; const struct snd_emu_chip_details *c; static struct snd_device_ops ops = { @@ -1867,6 +1900,8 @@ int snd_emu10k1_create(struct snd_card *card, is_audigy = emu->audigy = c->emu10k2_chip; + snd_emu10k1_detect_iommu(emu); + /* set addressing mode */ emu->address_mode = is_audigy ? 0 : 1; /* set the DMA transfer mask */ @@ -1893,8 +1928,13 @@ int snd_emu10k1_create(struct snd_card *card, emu->port = pci_resource_start(pci, 0); emu->max_cache_pages = max_cache_bytes >> PAGE_SHIFT; + + page_table_size = sizeof(u32) * (emu->address_mode ? MAXPAGES1 : + MAXPAGES0); + if (emu->iommu_workaround) + page_table_size += PAGE_SIZE; if (snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, snd_dma_pci_data(pci), - (emu->address_mode ? 32 : 16) * 1024, &emu->ptb_pages) < 0) { + page_table_size, &emu->ptb_pages) < 0) { err = -ENOMEM; goto error; } @@ -1910,8 +1950,11 @@ int snd_emu10k1_create(struct snd_card *card, goto error; } + silent_page_size = EMUPAGESIZE; + if (emu->iommu_workaround) + silent_page_size *= 2; if (snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, snd_dma_pci_data(pci), - EMUPAGESIZE, &emu->silent_page) < 0) { + silent_page_size, &emu->silent_page) < 0) { err = -ENOMEM; goto error; } @@ -1995,7 +2038,7 @@ int snd_emu10k1_create(struct snd_card *card, 0x00000000 | SPCS_EMPHASIS_NONE | SPCS_COPYRIGHT; /* Clear silent pages and set up pointers */ - memset(emu->silent_page.area, 0, PAGE_SIZE); + memset(emu->silent_page.area, 0, silent_page_size); silent_page = emu->silent_page.addr << emu->address_mode; for (idx = 0; idx < (emu->address_mode ? MAXPAGES1 : MAXPAGES0); idx++) ((u32 *)emu->ptb_pages.area)[idx] = cpu_to_le32(silent_page | idx); diff --git a/sound/pci/emu10k1/memory.c b/sound/pci/emu10k1/memory.c index a23eba540132..85414852c3db 100644 --- a/sound/pci/emu10k1/memory.c +++ b/sound/pci/emu10k1/memory.c @@ -457,6 +457,16 @@ static void get_single_page_range(struct snd_util_memhdr *hdr, *last_page_ret = last_page; } +static size_t synth_get_alloc_size(struct snd_emu10k1 *emu) +{ + size_t alloc_size = PAGE_SIZE; + + if (emu->iommu_workaround) + alloc_size *= 2; + + return alloc_size; +} + /* * allocate kernel pages */ @@ -471,7 +481,7 @@ static int synth_alloc_pages(struct snd_emu10k1 *emu, struct snd_emu10k1_memblk for (page = first_page; page <= last_page; page++) { if (snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, snd_dma_pci_data(emu->pci), - PAGE_SIZE, &dmab) < 0) + synth_get_alloc_size(emu), &dmab) < 0) goto __fail; if (!is_valid_page(emu, dmab.addr)) { snd_dma_free_pages(&dmab); @@ -488,7 +498,7 @@ static int synth_alloc_pages(struct snd_emu10k1 *emu, struct snd_emu10k1_memblk for (page = first_page; page <= last_page; page++) { dmab.area = emu->page_ptr_table[page]; dmab.addr = emu->page_addr_table[page]; - dmab.bytes = PAGE_SIZE; + dmab.bytes = synth_get_alloc_size(emu); snd_dma_free_pages(&dmab); emu->page_addr_table[page] = 0; emu->page_ptr_table[page] = NULL; @@ -513,7 +523,7 @@ static int synth_free_pages(struct snd_emu10k1 *emu, struct snd_emu10k1_memblk * continue; dmab.area = emu->page_ptr_table[page]; dmab.addr = emu->page_addr_table[page]; - dmab.bytes = PAGE_SIZE; + dmab.bytes = synth_get_alloc_size(emu); snd_dma_free_pages(&dmab); emu->page_addr_table[page] = 0; emu->page_ptr_table[page] = NULL;

1 0