]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
ALSA: usb: qcom: Fix false-positive address space check
authorTakashi Iwai <tiwai@suse.de>
Wed, 17 Sep 2025 13:09:01 +0000 (15:09 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 25 Sep 2025 09:16:52 +0000 (11:16 +0200)
commit09b473a80c1ce5e0d21680374b58d80278fbb1a4
tree36fa38c55dd26c58a0d2237d51f42ed868fdf6f3
parentb45cabfa717dd2fb1c0e85ad157f02513f576961
ALSA: usb: qcom: Fix false-positive address space check

[ Upstream commit 44499ecb4f2817743c37d861bdb3e95f37d3d9cd ]

The sanity check previously added to uaudio_transfer_buffer_setup()
assumed the allocated buffer being linear-mapped.  But the buffer
allocated via usb_alloc_coherent() isn't always so, rather to be used
with (SG-)DMA API.  This leaded to a false-positive warning and the
driver failed to work.

Actually uaudio_transfer_buffer_setup() deals only with the DMA-API
addresses for MEM_XFER_BUF type, while other callers of
uaudio_iommu_map() are with pages with physical addresses for
MEM_EVENT_RING and MEM_XFER_RING types.  So this patch splits the
mapping helper function to two different ones, uaudio_iommu_map() for
the DMA pages and uaudio_iommu_map_pa() for the latter, in order to
handle mapping differently for each type.  Along with it, the
unnecessary address check that caused probe error is dropped, too.

Fixes: 3335a1bbd624 ("ALSA: qc_audio_offload: try to reduce address space confusion")
Suggested-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Reported-and-tested-by: Luca Weiss <luca.weiss@fairphone.com>
Closes: https://lore.kernel.org/DBR2363A95M1.L9XBNC003490@fairphone.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
sound/usb/qcom/qc_audio_offload.c