As part of GSI reset workaround, a 1ms sleep needs to be added
between the first reset command and the second one.
Change-Id: I6131aa646edfd6192e4f0895a7cc97728d6f7536
CRs-Fixed: 1079245
Acked-by: Ady Abraham <adya@qti.qualcomm.com>
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
The range check for target load is incorrect. Fix this. This is only a
sanity check to catch badly specified target loads.
Change-Id: Ia90d020f5e0bdf37c600661a1c246dab5b637b3b
Signed-off-by: Olav Haugan <ohaugan@codeaurora.org>
bam can represent the transport used for the function. Hence change
the name from qcrndis to bam.
Change-Id: Ie153ae9cc00a72287b10623d63f29c08a1154cb9
Signed-off-by: Chandana Kishori Chiluveru <cchiluve@codeaurora.org>
The demux driver migration from kernel 3.18 to 4.4 added
extra parameter to dmx_ts_cb, dmx_section_cb callback functions
to send demux status to dmxdev filters. Revert back the change
to original prototype.
CRs-Fixed: 1079183
Change-Id: I4bdced9942e6524e8cc21410532fa7231049454f
Signed-off-by: Udaya Mallavarapu <udaym@codeaurora.org>
Add support for CDSP PIL which facilitates the loading of CDSP
firmware, authentication and bringing it out of reset.
Change-Id: I46b5da53614bdac740a1c85b0b3072cae32d20f7
Signed-off-by: Gaurav Kohli <gkohli@codeaurora.org>
It is possible that codec master clock enablement could race from two
different execution contexts, causing the mclk to be not enabled at all.
This will result in failure of use cases that expect the clock to be
present. Fix this issue by making sure the race condition does not
occur during mclk enablement.
bug: b/30983442
Change-Id: Ie254b8876524956b816267eaaed205f65641c000
Signed-off-by: Bhalchandra Gajare <gajare@codeaurora.org>
To help debugging, change adds support to read address and size from
the debugfs nodes and provide the memory content through the
mem_read debugfs node.
CRs-Fixed: 1064870
Change-Id: I151bc5d373129e3599baf6d3217b65843c0f16d0
Signed-off-by: Bhalchandra Gajare <gajare@codeaurora.org>
Update the power_supply framework of a dc insertion/removal.
Change-Id: I786d0774589ae64b298754286be28d67d394a66d
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
Current maximum transfer unit in the driver is set to 16K bytes,
whereas the maximum possible is upto 64K bytes. Update the driver
to use the maximum possible transfer unit and split the read and
write requests into chunks of this maximum possible transfer length.
CRs-Fixed: 1070087
Change-Id: I271cf0ad12652421ab3ae8770714ccb78c043efa
Signed-off-by: Bhalchandra Gajare <gajare@codeaurora.org>
During probe the driver does a devm_pinctrl_get() on the GPIOs assigned
to it. This call doesn't move the state of pins (bias/drv strength) to a
known state, rather only sets the function mux for those pins. As a result
the pins are in an unknown/floating state till a usecase for the port
begins.
Instead move the pins to the "sleep" state during probe to put them into a
known state.
Change-Id: Ifc6102d367fb0ca9fffbd50a6077dccd779e9a6c
Signed-off-by: Girish Mahadevan <girishm@codeaurora.org>
IPA RM keeps reference count for consumer resources.
In some cases where the client driver returns failure
on request_resource callback, the reference count may get
out of sync.
Change-Id: I6be84411e0bfe4f90570ab62c2d0a0c8539b8f34
CRs-Fixed: 1070261
Acked-by: Ady Abraham <adya@qti.qualcomm.com>
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
With change commit 0da8e4a21c ("icnss: Wait uninterruptible for
unregister driver"), return value changed from -ERESTARTSYS to -EINTR to
provide right return value to user space but as part of the change,
missed to update register driver case. Returning error even when driver
register is failed causes the module loader to remove the module because
of failure even though register driver as such is success.
CRs-fixed: 1078316
Change-Id: Ib3d13360def3ab47f121c66057c96977f5264043
Signed-off-by: Prashanth Bhatta <bhattap@codeaurora.org>
Correct the input argument to pass in the valid end address for the dmac
flush range function.
Change-Id: If6db015558a6a9e3db53ba7f1455808e779da908
Signed-off-by: Ram Chandrasekar <rkumbako@codeaurora.org>
Correct the input argument to pass in the valid end address for the dmac
flush range function.
Change-Id: Ib0e9690fc158a76dcebbd5ae45f67aaeca016a48
Signed-off-by: Ram Chandrasekar <rkumbako@codeaurora.org>
Correct the input argument to pass in the valid end address for the dmac
flush range function.
Change-Id: I2bc1eb26bcc7ed4aaa381417045d08b6779679ee
Signed-off-by: Ram Chandrasekar <rkumbako@codeaurora.org>
Correct the input argument to pass in the valid end address for the dmac
flush range function.
Change-Id: Iefcf85eaa5ea5542888269b7506b8f6e0e861243
Signed-off-by: Ram Chandrasekar <rkumbako@codeaurora.org>
Stop GSI channel for IPA producer endpoint includes
sending IPA DMA_TASK immediate command to IPA.
If the IPA DL group is in HOLB state, the DMA_TASK
will not be processed and ACK on it will not be sent to
the driver. In this case, ACK is redundant as the DL
data will release the IPA TX and GSI STOP indication
will be sent to S/W.
CRs-fixed: 1078380
Change-Id: I115524d562b63a8ec76b327207919b6ac9327fe2
Signed-off-by: Ghanim Fodi <gfodi@codeaurora.org>
