[alsa-devel] [PATCH 0/4] ASoC: Intel: Skylake: Fixes to free resources on init failure
Some resources are not freed for ipc and dsp init failures. This results in random failures in a stress test environment. So fixed to free the required resources and also fixed the DSP core reference counting for firmware load failure.
Subhransu S. Prusty (4): ASoC: Intel: Skylake: Fix to free dsp resource on ipc_init failure ASoC: Intel: Skylake: Fix to free resources for dsp_init failure ASoC: Intel: Skylake: Fix to free correct dev id in free_irq ASoC: Intel: Skylake: Fix DSP core ref count for init failure
sound/soc/intel/skylake/bxt-sst.c | 4 +++- sound/soc/intel/skylake/cnl-sst.c | 4 +++- sound/soc/intel/skylake/skl-messages.c | 26 +++++++++++++++++++------- sound/soc/intel/skylake/skl-sst-dsp.c | 4 ++-- sound/soc/intel/skylake/skl-sst.c | 4 +++- sound/soc/intel/skylake/skl.c | 2 +- 6 files changed, 31 insertions(+), 13 deletions(-)
For some dsp init error path, irq and few more resources are not freed. This results in oops. So, fix it by freeing up the resources on ipc_init failure.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com --- sound/soc/intel/skylake/bxt-sst.c | 4 +++- sound/soc/intel/skylake/cnl-sst.c | 4 +++- sound/soc/intel/skylake/skl-sst.c | 4 +++- 3 files changed, 9 insertions(+), 3 deletions(-)
diff --git a/sound/soc/intel/skylake/bxt-sst.c b/sound/soc/intel/skylake/bxt-sst.c index 71042aceb25f..88b8de342f31 100644 --- a/sound/soc/intel/skylake/bxt-sst.c +++ b/sound/soc/intel/skylake/bxt-sst.c @@ -582,8 +582,10 @@ int bxt_sst_dsp_init(struct device *dev, void __iomem *mmio_base, int irq, SKL_ADSP_W0_UP_SZ, BXT_ADSP_SRAM1_BASE, SKL_ADSP_W1_SZ);
ret = skl_ipc_init(dev, skl); - if (ret) + if (ret) { + skl_dsp_free(sst); return ret; + }
/* set the D0i3 check */ skl->ipc.ops.check_dsp_lp_on = skl_ipc_check_D0i0; diff --git a/sound/soc/intel/skylake/cnl-sst.c b/sound/soc/intel/skylake/cnl-sst.c index 7dcfe66a20e2..76ef1e40a50e 100644 --- a/sound/soc/intel/skylake/cnl-sst.c +++ b/sound/soc/intel/skylake/cnl-sst.c @@ -450,8 +450,10 @@ int cnl_sst_dsp_init(struct device *dev, void __iomem *mmio_base, int irq, CNL_ADSP_W1_SZ);
ret = cnl_ipc_init(dev, cnl); - if (ret) + if (ret) { + skl_dsp_free(sst); return ret; + }
cnl->boot_complete = false; init_waitqueue_head(&cnl->boot_wait); diff --git a/sound/soc/intel/skylake/skl-sst.c b/sound/soc/intel/skylake/skl-sst.c index 6cc586d3f03b..a6bcc0e6788f 100644 --- a/sound/soc/intel/skylake/skl-sst.c +++ b/sound/soc/intel/skylake/skl-sst.c @@ -562,8 +562,10 @@ int skl_sst_dsp_init(struct device *dev, void __iomem *mmio_base, int irq, SKL_ADSP_W0_UP_SZ, SKL_ADSP_SRAM1_BASE, SKL_ADSP_W1_SZ);
ret = skl_ipc_init(dev, skl); - if (ret) + if (ret) { + skl_dsp_free(sst); return ret; + }
sst->fw_ops = skl_fw_ops;
The patch
ASoC: Intel: Skylake: Fix to free dsp resource on ipc_init failure
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 3b3011adada3bba47c56c205634e1b32512e0c7c Mon Sep 17 00:00:00 2001
From: "Subhransu S. Prusty" subhransu.s.prusty@intel.com Date: Tue, 22 Aug 2017 16:45:50 +0530 Subject: [PATCH] ASoC: Intel: Skylake: Fix to free dsp resource on ipc_init failure
For some dsp init error path, irq and few more resources are not freed. This results in oops. So, fix it by freeing up the resources on ipc_init failure.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com Acked-By: Vinod Koul vinod.koul@intel.com Signed-off-by: Mark Brown broonie@kernel.org --- sound/soc/intel/skylake/bxt-sst.c | 4 +++- sound/soc/intel/skylake/cnl-sst.c | 4 +++- sound/soc/intel/skylake/skl-sst.c | 4 +++- 3 files changed, 9 insertions(+), 3 deletions(-)
diff --git a/sound/soc/intel/skylake/bxt-sst.c b/sound/soc/intel/skylake/bxt-sst.c index 5e1ac99ffa8b..4524211960e4 100644 --- a/sound/soc/intel/skylake/bxt-sst.c +++ b/sound/soc/intel/skylake/bxt-sst.c @@ -582,8 +582,10 @@ int bxt_sst_dsp_init(struct device *dev, void __iomem *mmio_base, int irq, SKL_ADSP_W0_UP_SZ, BXT_ADSP_SRAM1_BASE, SKL_ADSP_W1_SZ);
ret = skl_ipc_init(dev, skl); - if (ret) + if (ret) { + skl_dsp_free(sst); return ret; + }
/* set the D0i3 check */ skl->ipc.ops.check_dsp_lp_on = skl_ipc_check_D0i0; diff --git a/sound/soc/intel/skylake/cnl-sst.c b/sound/soc/intel/skylake/cnl-sst.c index 48ef04d5da89..387de388ce29 100644 --- a/sound/soc/intel/skylake/cnl-sst.c +++ b/sound/soc/intel/skylake/cnl-sst.c @@ -450,8 +450,10 @@ int cnl_sst_dsp_init(struct device *dev, void __iomem *mmio_base, int irq, CNL_ADSP_W1_SZ);
ret = cnl_ipc_init(dev, cnl); - if (ret) + if (ret) { + skl_dsp_free(sst); return ret; + }
cnl->boot_complete = false; init_waitqueue_head(&cnl->boot_wait); diff --git a/sound/soc/intel/skylake/skl-sst.c b/sound/soc/intel/skylake/skl-sst.c index a7ad734f591d..a436abf2fe3f 100644 --- a/sound/soc/intel/skylake/skl-sst.c +++ b/sound/soc/intel/skylake/skl-sst.c @@ -562,8 +562,10 @@ int skl_sst_dsp_init(struct device *dev, void __iomem *mmio_base, int irq, SKL_ADSP_W0_UP_SZ, SKL_ADSP_SRAM1_BASE, SKL_ADSP_W1_SZ);
ret = skl_ipc_init(dev, skl); - if (ret) + if (ret) { + skl_dsp_free(sst); return ret; + }
sst->fw_ops = skl_fw_ops;
unmap mmio and free memory resources if dsp_init fails.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com --- sound/soc/intel/skylake/skl-messages.c | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-)
diff --git a/sound/soc/intel/skylake/skl-messages.c b/sound/soc/intel/skylake/skl-messages.c index bfb3332a77ca..f0f11f597b21 100644 --- a/sound/soc/intel/skylake/skl-messages.c +++ b/sound/soc/intel/skylake/skl-messages.c @@ -277,8 +277,10 @@ int skl_init_dsp(struct skl *skl) }
ops = skl_get_dsp_ops(skl->pci->device); - if (!ops) - return -EIO; + if (!ops) { + goto unmap_mmio; + ret = -EIO; + }
loader_ops = ops->loader_ops(); ret = ops->init(bus->dev, mmio_base, irq, @@ -286,25 +288,35 @@ int skl_init_dsp(struct skl *skl) &skl->skl_sst);
if (ret < 0) - return ret; + goto unmap_mmio;
skl->skl_sst->dsp_ops = ops; cores = &skl->skl_sst->cores; cores->count = ops->num_cores;
cores->state = kcalloc(cores->count, sizeof(*cores->state), GFP_KERNEL); - if (!cores->state) - return -ENOMEM; + if (!cores->state) { + ret = -ENOMEM; + goto unmap_mmio; + }
cores->usage_count = kcalloc(cores->count, sizeof(*cores->usage_count), GFP_KERNEL); if (!cores->usage_count) { - kfree(cores->state); - return -ENOMEM; + ret = -ENOMEM; + goto free_core_state; }
dev_dbg(bus->dev, "dsp registration status=%d\n", ret);
+ return 0; + +free_core_state: + kfree(cores->state); + +unmap_mmio: + iounmap(mmio_base); + return ret; }
The patch
ASoC: Intel: Skylake: Fix to free resources for dsp_init failure
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 f77d443c4c299aff5ad9c74811dd063f4d8bebcf Mon Sep 17 00:00:00 2001
From: "Subhransu S. Prusty" subhransu.s.prusty@intel.com Date: Tue, 22 Aug 2017 16:45:51 +0530 Subject: [PATCH] ASoC: Intel: Skylake: Fix to free resources for dsp_init failure
unmap mmio and free memory resources if dsp_init fails.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com Acked-By: Vinod Koul vinod.koul@intel.com Signed-off-by: Mark Brown broonie@kernel.org --- sound/soc/intel/skylake/skl-messages.c | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-)
diff --git a/sound/soc/intel/skylake/skl-messages.c b/sound/soc/intel/skylake/skl-messages.c index bfb3332a77ca..f0f11f597b21 100644 --- a/sound/soc/intel/skylake/skl-messages.c +++ b/sound/soc/intel/skylake/skl-messages.c @@ -277,8 +277,10 @@ int skl_init_dsp(struct skl *skl) }
ops = skl_get_dsp_ops(skl->pci->device); - if (!ops) - return -EIO; + if (!ops) { + goto unmap_mmio; + ret = -EIO; + }
loader_ops = ops->loader_ops(); ret = ops->init(bus->dev, mmio_base, irq, @@ -286,25 +288,35 @@ int skl_init_dsp(struct skl *skl) &skl->skl_sst);
if (ret < 0) - return ret; + goto unmap_mmio;
skl->skl_sst->dsp_ops = ops; cores = &skl->skl_sst->cores; cores->count = ops->num_cores;
cores->state = kcalloc(cores->count, sizeof(*cores->state), GFP_KERNEL); - if (!cores->state) - return -ENOMEM; + if (!cores->state) { + ret = -ENOMEM; + goto unmap_mmio; + }
cores->usage_count = kcalloc(cores->count, sizeof(*cores->usage_count), GFP_KERNEL); if (!cores->usage_count) { - kfree(cores->state); - return -ENOMEM; + ret = -ENOMEM; + goto free_core_state; }
dev_dbg(bus->dev, "dsp registration status=%d\n", ret);
+ return 0; + +free_core_state: + kfree(cores->state); + +unmap_mmio: + iounmap(mmio_base); + return ret; }
The dev_id passed by the driver in request_threaded_irq is an ebus pointer, whereas to free_irq it is hdac_bus. Fix by passing correct dev_id to free_irq.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com --- sound/soc/intel/skylake/skl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/sound/soc/intel/skylake/skl.c b/sound/soc/intel/skylake/skl.c index e25e5ed464dc..f94b484abb99 100644 --- a/sound/soc/intel/skylake/skl.c +++ b/sound/soc/intel/skylake/skl.c @@ -415,7 +415,7 @@ static int skl_free(struct hdac_ext_bus *ebus) snd_hdac_ext_stop_streams(ebus);
if (bus->irq >= 0) - free_irq(bus->irq, (void *)bus); + free_irq(bus->irq, (void *)ebus); snd_hdac_bus_free_stream_pages(bus); snd_hdac_stream_free_all(ebus); snd_hdac_link_free_all(ebus);
The patch
ASoC: Intel: Skylake: Fix to free correct dev id in free_irq
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 c360e0c3ab8ef8d0fd6b5501a407b9a2be4f204d Mon Sep 17 00:00:00 2001
From: "Subhransu S. Prusty" subhransu.s.prusty@intel.com Date: Tue, 22 Aug 2017 16:45:52 +0530 Subject: [PATCH] ASoC: Intel: Skylake: Fix to free correct dev id in free_irq
The dev_id passed by the driver in request_threaded_irq is an ebus pointer, whereas to free_irq it is hdac_bus. Fix by passing correct dev_id to free_irq.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com Acked-By: Vinod Koul vinod.koul@intel.com Signed-off-by: Mark Brown broonie@kernel.org --- sound/soc/intel/skylake/skl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/sound/soc/intel/skylake/skl.c b/sound/soc/intel/skylake/skl.c index e612a4097275..d62ffbc93d54 100644 --- a/sound/soc/intel/skylake/skl.c +++ b/sound/soc/intel/skylake/skl.c @@ -415,7 +415,7 @@ static int skl_free(struct hdac_ext_bus *ebus) snd_hdac_ext_stop_streams(ebus);
if (bus->irq >= 0) - free_irq(bus->irq, (void *)bus); + free_irq(bus->irq, (void *)ebus); snd_hdac_bus_free_stream_pages(bus); snd_hdac_stream_free_all(ebus); snd_hdac_link_free_all(ebus);
During dsp init failure, the ref count is not incremented and dsp is powered down. But as the skl driver calls put_core for the init failure it decrements the dsp core ref count and ref count becomes unbalanced.
This results in dsp core powered up in further runtime suspend/resume cycles and never powered down.
So increment the ref count before dsp core powerup and for any failure, decrement in put_core will be balanced.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com --- sound/soc/intel/skylake/skl-sst-dsp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/sound/soc/intel/skylake/skl-sst-dsp.c b/sound/soc/intel/skylake/skl-sst-dsp.c index c2ba5ff1632c..19ee1d4f3bdf 100644 --- a/sound/soc/intel/skylake/skl-sst-dsp.c +++ b/sound/soc/intel/skylake/skl-sst-dsp.c @@ -351,6 +351,8 @@ int skl_dsp_get_core(struct sst_dsp *ctx, unsigned int core_id) return -EINVAL; }
+ skl->cores.usage_count[core_id]++; + if (skl->cores.state[core_id] == SKL_DSP_RESET) { ret = ctx->fw_ops.set_state_D0(ctx, core_id); if (ret < 0) { @@ -359,8 +361,6 @@ int skl_dsp_get_core(struct sst_dsp *ctx, unsigned int core_id) } }
- skl->cores.usage_count[core_id]++; - out: dev_dbg(ctx->dev, "core id %d state %d usage_count %d\n", core_id, skl->cores.state[core_id],
The patch
ASoC: Intel: Skylake: Fix DSP core ref count for init failure
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 7b992c24de78a206b3abd07192686d2d5db5012c Mon Sep 17 00:00:00 2001
From: "Subhransu S. Prusty" subhransu.s.prusty@intel.com Date: Tue, 22 Aug 2017 16:45:53 +0530 Subject: [PATCH] ASoC: Intel: Skylake: Fix DSP core ref count for init failure
During dsp init failure, the ref count is not incremented and dsp is powered down. But as the skl driver calls put_core for the init failure it decrements the dsp core ref count and ref count becomes unbalanced.
This results in dsp core powered up in further runtime suspend/resume cycles and never powered down.
So increment the ref count before dsp core powerup and for any failure, decrement in put_core will be balanced.
Signed-off-by: Subhransu S. Prusty subhransu.s.prusty@intel.com Acked-By: Vinod Koul vinod.koul@intel.com Signed-off-by: Mark Brown broonie@kernel.org --- sound/soc/intel/skylake/skl-sst-dsp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/sound/soc/intel/skylake/skl-sst-dsp.c b/sound/soc/intel/skylake/skl-sst-dsp.c index c2ba5ff1632c..19ee1d4f3bdf 100644 --- a/sound/soc/intel/skylake/skl-sst-dsp.c +++ b/sound/soc/intel/skylake/skl-sst-dsp.c @@ -351,6 +351,8 @@ int skl_dsp_get_core(struct sst_dsp *ctx, unsigned int core_id) return -EINVAL; }
+ skl->cores.usage_count[core_id]++; + if (skl->cores.state[core_id] == SKL_DSP_RESET) { ret = ctx->fw_ops.set_state_D0(ctx, core_id); if (ret < 0) { @@ -359,8 +361,6 @@ int skl_dsp_get_core(struct sst_dsp *ctx, unsigned int core_id) } }
- skl->cores.usage_count[core_id]++; - out: dev_dbg(ctx->dev, "core id %d state %d usage_count %d\n", core_id, skl->cores.state[core_id],
On Tue, Aug 22, 2017 at 04:45:49PM +0530, Subhransu S. Prusty wrote:
Some resources are not freed for ipc and dsp init failures. This results in random failures in a stress test environment. So fixed to free the required resources and also fixed the DSP core reference counting for firmware load failure.
Acked-By: Vinod Koul vinod.koul@intel.com
Subhransu S. Prusty (4): ASoC: Intel: Skylake: Fix to free dsp resource on ipc_init failure ASoC: Intel: Skylake: Fix to free resources for dsp_init failure ASoC: Intel: Skylake: Fix to free correct dev id in free_irq ASoC: Intel: Skylake: Fix DSP core ref count for init failure
sound/soc/intel/skylake/bxt-sst.c | 4 +++- sound/soc/intel/skylake/cnl-sst.c | 4 +++- sound/soc/intel/skylake/skl-messages.c | 26 +++++++++++++++++++------- sound/soc/intel/skylake/skl-sst-dsp.c | 4 ++-- sound/soc/intel/skylake/skl-sst.c | 4 +++- sound/soc/intel/skylake/skl.c | 2 +- 6 files changed, 31 insertions(+), 13 deletions(-)
-- 1.9.1
participants (3)
-
Mark Brown
-
Subhransu S. Prusty
-
Vinod Koul