Use spin_lock_irqsave utility to avoid preemtpion of process_cmd thread
in jpeg. This can come in rare conditions when IRQ is received on same
CPU which is handling the thread.
Such cases will result in deadlock conditions.
CRs-Fixed: 3712176
Change-Id: Ib0d9997495272ab49a75b0419eb1e0fd84b1962c
Signed-off-by: Dharmender Sharma <quic_dharshar@quicinc.com>
Use bubble packet for request reapply for all scenarios
irrespective of mode switch. Bubble packet carries exposure
updates as well.
CRs-Fixed: 3689201
Change-Id: If2170fd77ea7fe88155ed7101fc053d13bd592a8
Signed-off-by: Karthik Anantha Ram <quic_kartanan@quicinc.com>
(cherry picked from commit c0b8a1bad215c6ba71851897ad642611f085dfb3)
Fix buf done handling in tfe bus driver due to missing
handling for multiple out resource.
CRs-Fixed: 3696470
Signed-off-by: Alok Chauhan <quic_alokc@quicinc.com>
Change-Id: Ie5597b83d21368f921d232d242c82be671ab8080
Output port map structure is not defined for TFE.
This is causing null pointer dereference in add io
buffer api.
Add support for outport map structure in TFE driver.
CRs-Fixed: 3696470
Signed-off-by: Alok Chauhan <quic_alokc@quicinc.com>
Change-Id: I35fa1f36a65bb0ba3dfd896567b47549b6bebf9d
Flush all cmd release all request from CRE request list, once release
these request should not be access by other threads. This commit
add proper mutex lock and conditions to prevent use of released request
from CRE request list. Also change SW reset to HW reset in case of
flush all.
CRs-Fixed: 3583508
Change-Id: I35b38bf5b1771a26bb5a37d3404f6e577abcb66b
Signed-off-by: Ayush Kumar <quic_ayushkr@quicinc.com>
Mutex unlock was missing in CSID driver in case the drv clk rate
call fails. This can result in recursive locking situations.
This commit handles the error scenario.
Change-Id: Ie1e60006713f0a96831ec1174bec1b40ab460a1f
CRs-Fixed: 3712878
Signed-off-by: Gaurav Jindal <quic_gjindal@quicinc.com>
Trigger callback to ISP each time when there is a change
in phy clock so that csid and tfe clock rate can be
updated accordingly. The argument passed in callback api is
the final phy clock rate that is applied.
CRs-Fixed: 2854066
Change-Id: Ia74465e4082e08b4c50338e0eb779f996afa2795
Signed-off-by: shiwgupt <quic_shiwgupt@quicinc.com>
Signed-off-by: shaduls <quic_shaduls@quicinc.com>
(cherry picked from commit 1fd5054f530a16a5bf90901ba4c8bf2025e136dc)
Program WM mode based on mode flag for RDI resources..
CRs-Fixed: 3627088
Change-Id: Ie2918462fdfa8fe486983e63fcf66f9ec568c75f
Signed-off-by: Chandan Kumar Jha <quic_cjha@quicinc.com>
Enhance support for HW capability for other targets.
CRs-Fixed: 3696470
Signed-off-by: Alok Chauhan <quic_alokc@quicinc.com>
Change-Id: I53d20dda4cad7cb653808d7e93911297979dce25
Distinguish error notification to CRM for actual bubble &
other errors. This avoids conflict in internal recovery scheme.
Currently back to back bubble triggers an internal recovery, if
an error is notified for AEB back to back it will lead to two
different callflows for internal recovery. Distinguishing the
notifications will avoid different callflows.
CRs-Fixed: 3692264
Change-Id: Id4d928569fded732cef1782863ee494c66158c44
Signed-off-by: Karthik Anantha Ram <quic_kartanan@quicinc.com>
(cherry picked from commit a0a0290f13ddb1adb25eee5e55ee4f438524fbd9)
adding missing call for put buf in flash to avoid leak.
CRs-Fixed: 3666659
Change-Id: I2e08331654e63b77950bb83ee6d95c4f5a716a86
Signed-off-by: jinguiw <quic_jinguiw@quicinc.com>
(cherry picked from commit 92e992dfae3cd269fbba0b51f0b456bebbb5f797)
Detecting for internal recovery and flush execute in two
different thread contexts, and the critical section protecting
the lists is small to avoid locking overhead. It is possible that flush
was triggered, but bh in ISP detected an error condition and triggers
a internal recovery moving the request in question to pending list.
At this point it is possible, that the flush context has cleared
the pending list, and internal recovery's addition to the pending list
is agnostic to the flushing thread. When there is a flush all, a reset
is issued to the HW and all the bh's are flushed as well. Post this
clearing the pending request list ensures no stale request entity is
left behind as opposed to cleaning up the pending list prior to the
HW stop sequence.
CRs-Fixed: 3669530
Change-Id: I58f024ec51cbc9c391d8aab1199098f314317b8a
Signed-off-by: Karthik Anantha Ram <quic_kartanan@quicinc.com>
(cherry picked from commit 0d58af6cf25db3b3412bc6effc437474751bf700)
Increase total link capacity, but cap session capacity at 4 links.
CRs-Fixed: 3663731
Change-Id: I53e545cb52a89fb15a254426653648a1b7b6226f
Signed-off-by: Karthik Anantha Ram <quic_kartanan@quicinc.com>
(cherry picked from commit cc952af03fbad9a749f861dbb14738e9fbc34a56)
Query the context bus comp group information in legacy acquire
call.
CRs-Fixed: 3696470
Signed-off-by: Alok Chauhan <quic_alokc@quicinc.com>
Change-Id: Ib80858e6b81e1c7fce2e00597f58f0a98e12ff87
Removed the camera subdev notify message at top half and
send subdev notification in the bottom half.
CRs-Fixed: 3645184
Change-Id: Iee60b243c9ca1241533dfe85749198b3f187618f
Signed-off-by: Sourabh Soni <quic_soursoni@quicinc.com>
For tfe Out resource data, iterate over num_out count instead of
TFE_OUT_MAX.
CRs-Fixed: 3604707
Change-Id: I63ea618da041bbf618c8bdc373daa13820f68c73
Signed-off-by: Pranav Sanwal <quic_psanwal@quicinc.com>
(cherry picked from commit d4fb18dc1347ff1e49136cdab1f0557dd86fb657)
Stale entry in TFE WM Clients data adds some offset in image address,
due to this TFE page fault observed. This commit adds change to reset
flags and entries of TFE WM Clients related to buffer alignment.
CRs-Fixed: 3569926
Change-Id: I5ac48cc2b0855f008b05c23a8c4645aeae348c6b
Signed-off-by: Ayush Kumar <quic_ayushkr@quicinc.com>
