Skip to contentSkip to conversation

ALSA: pcm: Fix mmap capability check

PATCH In review Takashi Iwai 2 participants 3 messages Jul 20, 2021 – Jul 30, 2021 Sound

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ALSA: pcm: Fix mmap capability check

The hw_support_mmap() doesn't cover all memory allocation types and might use a wrong device pointer for checking the capability. Check the all memory allocation types more completely.

---
 sound/core/pcm_native.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
sound/core/pcm_native.c +10 -4
index c88c4316c417..6919d2943b9d 100644
@@ -246,12 +246,18 @@ static bool hw_support_mmap(struct snd_pcm_substream *substream)
 	if (!(substream->runtime->hw.info & SNDRV_PCM_INFO_MMAP))
 		return false;
 
-	if (substream->ops->mmap ||
-	    (substream->dma_buffer.dev.type != SNDRV_DMA_TYPE_DEV &&
-	     substream->dma_buffer.dev.type != SNDRV_DMA_TYPE_DEV_UC))
+	if (substream->ops->mmap)
 		return true;
 
-	return dma_can_mmap(substream->dma_buffer.dev.dev);
+	switch (substream->dma_buffer.dev.type) {
+	case SNDRV_DMA_TYPE_UNKNOWN:
+		return false;
+	case SNDRV_DMA_TYPE_CONTINUOUS:
+	case SNDRV_DMA_TYPE_VMALLOC:
+		return true;
+	default:
+		return dma_can_mmap(substream->dma_buffer.dev.dev);
+	}
 }
 
 static int constrain_mask_params(struct snd_pcm_substream *substream,
Signature · 1 line
2.26.2
maintainer
#1 lore raw
Quoted 3 lines · Takashi Iwai

The hw_support_mmap() doesn't cover all memory allocation types and might use a wrong device pointer for checking the capability. Check the all memory allocation types more completely.

This change breaks mmap for the snd-dummy driver (fake_buffer). It seems that we need this fix?

sound/core/pcm_native.c +0 -0
index 6a2971a7e6a1..09c0e2a6489c 100644
@@ -246,7 +246,7 @@ static bool hw_support_mmap(struct snd_pcm_substream

*substream)

        if (!(substream->runtime->hw.info & SNDRV_PCM_INFO_MMAP))
                return false;

- if (substream->ops->mmap)
+ if (substream->ops->mmap || substream->ops->page)

                return true;

        switch (substream->dma_buffer.dev.type) {


					Jaroslav
Quoted 35 lines
---
 sound/core/pcm_native.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
sound/core/pcm_native.c +10 -4
index c88c4316c417..6919d2943b9d 100644
@@ -246,12 +246,18 @@ static bool hw_support_mmap(struct snd_pcm_substream *substream)
 	if (!(substream->runtime->hw.info & SNDRV_PCM_INFO_MMAP))
 		return false;
 
-	if (substream->ops->mmap ||
-	    (substream->dma_buffer.dev.type != SNDRV_DMA_TYPE_DEV &&
-	     substream->dma_buffer.dev.type != SNDRV_DMA_TYPE_DEV_UC))
+	if (substream->ops->mmap)
 		return true;
 
-	return dma_can_mmap(substream->dma_buffer.dev.dev);
+	switch (substream->dma_buffer.dev.type) {
+	case SNDRV_DMA_TYPE_UNKNOWN:
+		return false;
+	case SNDRV_DMA_TYPE_CONTINUOUS:
+	case SNDRV_DMA_TYPE_VMALLOC:
+		return true;
+	default:
+		return dma_can_mmap(substream->dma_buffer.dev.dev);
+	}
 }
 
 static int constrain_mask_params(struct snd_pcm_substream *substream,
Signature · 2 lines
Jaroslav Kysela <[email protected]>
Linux Sound Maintainer; ALSA Project; Red Hat, Inc.
#2 lore raw
Quoted 22 lines · Jaroslav Kysela
Quoted 3 lines · Takashi Iwai

The hw_support_mmap() doesn't cover all memory allocation types and might use a wrong device pointer for checking the capability. Check the all memory allocation types more completely.

This change breaks mmap for the snd-dummy driver (fake_buffer). It seems that we need this fix?

sound/core/pcm_native.c +1 -1
index 6a2971a7e6a1..09c0e2a6489c 100644
@@ -246,7 +246,7 @@ static bool hw_support_mmap(struct snd_pcm_substream
 *substream)
        if (!(substream->runtime->hw.info & SNDRV_PCM_INFO_MMAP))
                return false;
 
-       if (substream->ops->mmap)
+       if (substream->ops->mmap || substream->ops->page)
                return true;
 
Quoted 1 line
        switch (substream->dma_buffer.dev.type) {

Yes, that would work. Or allowing the mmap for unknown buffer type by blindly relying to the hw.info flag. But ops->page check looks safer.

Care to submit a proper patch? Thanks!

Takashi