Uplink mute observed in voice call when Vbat is enabled. This is due
to external echo reference port "SLIMBUS_1_TX" is not started, as it
is not connected to any of the voice front end DAI links. Fix this
issue by updating dapm route map b/w SLIMBUS_1_TX and voice front end
DAI links.
Change-Id: I02ffbb2565deed073909d49b3fb20548bc406daa
Signed-off-by: Venkata Narendra Kumar Gutta <vgutta@codeaurora.org>
Clean up the client pid information from the client table on
lsm_deinit Ioctl call.
Change-Id: Ia40a481b1be909c46371c0bb295db7baa30bdbc8
Signed-off-by: Sreelakshmi Gownipalli <sgownipa@codeaurora.org>
In CPP driver, rate limit get_buf error messages
for POP_STREAM IOCTL because this IOCTL can be triggered
too often (in some error conditions) and error messages here
can cause watchdog issues.
Change-Id: Idd0428ac9f1b003c5d080ded0b06836e8500325a
Signed-off-by: Krishna Srinivas Kundurthi <krisri@codeaurora.org>
Enable msm boot stats for MSM8952 by adding the MPM
sleep counter node, which is needed to report msm
boot stats such as bootloader throughput, total boot
time, etc.
Change-Id: Ia81ac6af086cb87f43e10e80aa8dacb78bbbb45a
Signed-off-by: Kishan Kumar <kishank@codeaurora.org>
Send cluster pm notifications on core,L2 and system collapse.
Change-Id: I24c9e97440fe45e00f47bb00839194a2177d4c66
Signed-off-by: Achyuth Sai Vadrav <avadra@codeaurora.org>
If the audio mixing is done within codec, then echo reference
needs to be obtained from the mixed output in the codec.
Add support to get the external echo reference from SLIMBUS_1_TX
afe port.
Change-Id: I046172c02dc2eac5d1b8156abca3e4ca960d78f4
Signed-off-by: Phani Kumar Uppalapati <phaniu@codeaurora.org>
In the case when subsystem restart is triggered on a subsystem
which is being shutdown, there might be a race condition which
could involve the SSR framework trying to restart a subsystem
which is already shutdown. Print a warning and return from the
SSR function if this is the case.
Change-Id: I38083822db598e7b292bcf3b9e187a9f56d48bbb
Signed-off-by: Deepak Katragadda <dkatraga@codeaurora.org>
Add clk ctrl configs for cursor pipes. These are used to
check the status of cursor pipes.
Change-Id: Idfb88b8ba488bc563e4f9393ca25145334d6d60f
Signed-off-by: Jayant Shekhar <jshekhar@codeaurora.org>
Source crop for cursor is now sent along with the image width/height.
Previously, cursor was always set with the image with/height as there
was no crop information. And allocate and map/unmap cursor buffer for
every cursor enable/disable. This would address the issue of cursor
image being more than the supported cursor size, as previously cursor
buffer was alloted only to a fixed cursor size.
Change-Id: If9bdd6887a9f0ab33751f34583e5eb671a19f9fa
Signed-off-by: Veera Sundaram Sankaran <veeras@codeaurora.org>
The clk status register and vbif status based on xin-id
is used to check if pipes are idle. VBIF checks cannot be done
for cursor pipes as the same xin-id is shared between dsi, cursor0
and cursor1. Avoid vbif status check and fix clk status check for
cursor pipes.
Change-Id: I9e0a185beffd4e732c1b5dc61822cc94b3735a27
Signed-off-by: Veera Sundaram Sankaran <veeras@codeaurora.org>
Cursor IOCTL is used to program the cursor pipes and the updates
would take effect on display_commit along with the other layers.
In cmd mode panels, when the there are no updates there is idle
power collapse and iommu is detached. If cursor IOCTL is called
during this time, it would assign the physical address instead of
virtual address for cursor buffer. This causes dirty region around
the cursor in the next display_commit. Fix the cursor buffer to have
the virtual address all the time.
Change-Id: I963794d0de5387b10e06e7c5a61300cd7b4aeffa
Signed-off-by: Veera Sundaram Sankaran <veeras@codeaurora.org>
commit: cc4d47ad41
msm: mdss: add async support for hw cursor, introduced the async
support for cursor. To support this, both left and right mixer
of the cursor pipe is staged for dual mixer targets. When the first
cursor update falls on the right LM, only the right mixer is staged,
causing underrun when the cursor position moves to left LM with async
updates. Fix staging of both LM for cursor pipes.
Change-Id: If6f5a35ee1e9456d93a87d150e6a60452b127b2b
Signed-off-by: Veera Sundaram Sankaran <veeras@codeaurora.org>
Pipes are marked dirty when the userspace and kernel pipe management
are out of sync and the kernel stabilizes itself with the next prepare
call. The prepare call would not involve cursor pipes, as they are
handled separately through the cursor ioctl. This leaves the cursor
pipe to stay dirty and fails the following pre_commit. Fix to avoid
setting cursor pipe as dirty.
Change-Id: I7ba6347290e47cb04127da8ae82e41f0e9f5d65f
Signed-off-by: Veera Sundaram Sankaran <veeras@codeaurora.org>
BAM2BAM mode is not designed to work in Full speed mode and USB HW
behaviour is unknown in case request is queued to BAM2BAM endpoint.
Hence Fail request queueing for BAM endpoints in Full speed mode
as per recommendation from hardware programming guide.
CRs-Fixed: 884453
Change-Id: I29bcdde47bda63c5de15ef8151189888c9cda8da
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
