In QC platform, PCIe device is enumerated when system boot up,
cnss platform driver should reject wlan driver whose supported
PCIe device isn't existing.
CRs-Fixed: 2647082
Change-Id: Ie6579d1ed34de2c183f9ce2d411a02d6c8137da5
Signed-off-by: Kai Liu <kaliu@codeaurora.org>
In some conditions tasklet was exiting before processing
the all the packets. Add changes to restart tasklet for
pending packets.
Change-Id: I7246c070a003e27e8ef0b7036cf275a6f072b1b1
Signed-off-by: Ashok Vuyyuru <avuyyuru@codeaurora.org>
From 5.3+ kernels we have an option that users of the dmabuf can
provide unique name to the dmabuf object which is stored in the ->name
field of dma_buf structure. The change, commit bf93f26d49 ("UPSTREAM:
dma-buf: add DMA_BUF_SET_NAME ioctls"), brought as an LSK merges can
break some users. These users, say display, relies on the readlink() to
get the dmabuf name which is used to uniquely identify the dmabuf
exported buffers and thus take some decissions, Eg: Whether to redraw
the eglimage. But this LSK change will always give the "/dmabuf:" as
name for all the dmabuf objects thus breaking such users. This can be
avoided by filling the dmabuf->name with the default unique name, thus
the dma_buf->buf_name,name will share the common name till the user
overrides it.
This fix also indirectly solves the memory leak issue associated with
the dmabuf->buf_name filled in the dma_buf_export().
Change-Id: If4ba945b49c98f1d109ec83400299af12dac1232
Fixes: bf93f26d49 ("UPSTREAM: dma-buf: add DMA_BUF_SET_NAME ioctls")
Signed-off-by: Charan Teja Reddy <charante@codeaurora.org>
Existing scan logic will start repetitive seek
from lower freq and stops when band limit reached,
with this logic valid lower band is always ignored.
Seek one more time one band limit reached to check
if lower freq channel is valid.
CRs-Fixed: 2683848
Change-Id: I0b2608b544e0402298d1dfdc1bef96ff8ca64123
Signed-off-by: himta ram <hram@codeaurora.org>
If FM is tuned to invalid freq and scan performed then invalid
freq also included in scan list. Post the tune event with
original frequency after the scan complete event posted to
user space to avoid the invalid tuned freq in scan list.
CRs-Fixed: 2681008
Change-Id: I8ac3b43742a4f92272fc5b0283fabcad2e04de5b
Signed-off-by: himta ram <hram@codeaurora.org>
Since idle restart is a multiple stages power up sequence, there will be
race conditions when idle restart and unregister driver (which is power
down sequence) come at the same time. Add mutex protection for them so
that they can be serialized completely.
Change-Id: Icc28b5f759cd690cb505b16ab70eaa4abaf2e4cc
Signed-off-by: Yue Ma <yuem@codeaurora.org>
Add extra checks to avoid NULL pointer derefernce when DEBUG_FS
is disabled.
Change-Id: I46ff3b8b27b53ff5bb5dead78ccdefc410435dda
Acked-by: Tadakamalla Krishnaiah <ktadakam@qti.qualcomm.com>
Signed-off-by: Mohammed Nayeem Ur Rahman <mohara@codeaurora.org>
Enable the Hardware key manager driver that interfaces
with the master and slave blocks for crypto functionality.
Change-Id: Ie8492bf80e357ae1ddc957728e121f49d79ae61e
Signed-off-by: AnilKumar Chimata <anilc@codeaurora.org>
EUD may be enabled in boot loader and to keep EUD session
alive across kernel boot till USB phy driver is initialized
based on cable status, keep regulators on from USB high speed
phy driver's probe. Regulators shall get turn off while
handling USB cable disconnect routine for both boot up with
and without USB cable case.
Change-Id: I7ebdc0237d7db3e45b61a57264a5564bfa01ed1c
Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org>
The PMIC firmware is no more enabling the resources required
for the PHY for the EUD usecase.
So, if the required resources for the PHY are turned off
as part of spoof disconnect the EUD fails to work.
Fix this by keeping the PHY resources enabled as part
of spoof disconnect.
Change-Id: I2ee734eec0c78b062d2e0726e9efa5a88db8c02d
Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org>
Update the name of the sysfs node to extended_cache_enable
for proper readability.
Change-Id: I11e16d7df72e863547dc6880b64a94fbca1c7413
Signed-off-by: Raghavendra Kakarla <rkakarla@codeaurora.org>
First, it should be noted that the CQE timeout (60 seconds) is
substantial so a CQE request that times out is really stuck, and the
race between timeout and completion is extremely unlikely. Nevertheless
this patch fixes an issue with it.
Commit ad73d6fead ("mmc: complete requests from ->timeout")
preserved the existing functionality, to complete the request.
However that had only been necessary because the block layer
timeout handler had been marking the request to prevent it from being
completed normally. That restriction was removed at the same time, the
result being that a request that has gone will have been completed
anyway.
That is, the completion was unnecessary.
At the time, the unnecessary completion was harmless because the block
layer would ignore it, although that changed in kernel v5.0.
Change-Id: I8690c293afa280bb030d114008fb00df711417f8
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Fixes: ad73d6fead ("mmc: complete requests from ->timeout")
Cc: stable@vger.kernel.org
Git-commit: c077dc5e0620508a29497dac63a2822324ece52a
Git-repo: https://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc.git
Link: https://lore.kernel.org/r/20200508062227.23144-1-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Sarthak Garg <sartgarg@codeaurora.org>
