| Commit message (Collapse) | Author | Age |
|---|
| ... | |
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The commit "qcacld: Defer vdev deletion until del_bss is completed"
fixes the race condition observed in vdev removal during p2p
operation. But the same problem also observed with other modes as
well. Since bss_status is being accessing in different contexts,
changing it into atomic type and also STOPPED status should be set
only after flushing the pending frames in vdev. As of now, STARTED
status is set in two places (add_sta for STA mode, add_bss for
AP/IBSS mode).
Change-Id: I3590d14f6ded44d44ec647f9d21ee2e1e2f76010
CRs-Fixed: 610490
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
- Before accessing the interrupt enable and clear registers,
wake the SoC in interrupt handler and in tasklet
- Disable IRQs as well while holding spinlock
when waking or putting target to sleep to avoid deadlock.
- Add debug when interrupt cannot be disabled due to loss of
synchronization between host and target states during suspend
Change-Id: I24c1b607e5680bd5a56141b332f705e7705fc54e
CRs-Fixed: 600998
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The regulatory callback can get called while the driver is getting
unloaded, leading to race condition. Check for the load-unload flag
to prevent this race condition.
Change-Id: I68645ee24d0a85fc50b19223f0efddde77c334ac
CRs-Fixed: 610480
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
It is possible to run into an infinite loop of CE Rx when
doing more Rx completion. Have a threshold count (20) for
more completion. Do not do anymore more Rx completion once
we hit this threshold value.
Change-Id: I35a892f90400f2462035d12ddf87ded4b4ba1d4c
CRs-Fixed: 610472
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Prevent suspend and vote for QoS latency when system
is in SAP/P2PGO mode with clients.
Allow suspend and remove QoS request when no client
is connected.
Change-Id: I15136585e68232b30b612d67c2a8358bad0d0260
CRs-Fixed: 605324
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
During LFR_MCC testing, we need to enable DEBUG level logs to debug
some of the issues, but with the logs in MCC tx pause handler we are
seeing crash due to watchdog kicking in due to too many logs as this
function is called in high intensity pause/unpause events.
The fix is to:
1. Move mcc tx pause handler log statements to INFO level
2. Also move some of the other logs to info level
3. Remove \n at the end of WMA logs which are unnecessary and fills
kernel log buffer
Change-Id: I03c41b80672aebc550f2b5d7cbe8c62f9ce421d4
CRs-Fixed: 610735
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
force_break flag is not reset anywhere. Having
stale value might limit the prcoessing of Rx
messages. Also make sure to reset force_break
flags out side of more_completions label
to fix any corner case which could result in
processing more than MAX_NUM_OF_RECEIVES Rx
frames in single intr handling by doing more
completions.
Change-Id: I3b7ef12e2be38ee67d0c657d5bc47d7031498031
CRs-Fixed: 610531
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Host must not configure vdev id greater than the maximum number of
interfaces acceptable by the firmware. In the current design, there is
control for interfaces created from the userspace. But when vdev id is
corrupted (for e.g., in error cases due to improper cleanup), the host
must still ensure that the vdev id configured to the firwmare is
actually valid
Change-Id: I63e99e7300d7f05ed1490f14d8711f6c56f759f7
CRs-Fixed: 610642
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Both wma_stop() and wma_suspend_target() prints the same error message
"Failed to suspend target" when firmware fails to send ACK to the host
for WMI_PDEV_SUSPEND_CMDID. Getting "Failed to suspend target" message
twice in dmesg creates confusion to the user.
Change-Id: Id9efda83589e872797cd5ddad2f8c6545f014f77
CRs-Fixed: 610565
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Added debug log to display TX power
sent to firmware for all channels in
update channel list WMI command.
Change-Id: I6498fe2374ef05644a815865a58b7967099bfe90
CRs-Fixed: 591370
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The root cause of the issue is during roaming we do netif_tx_disable
to stall the Tx data and host issues ADD_BSS_REQ (vdev_start) for
which FW is not sending any response and vdev_response timer times out.
Upon timeout, The code to post the TIMEOUT message back to LIM is
not written; hence Lim is blocked waiting for ADD_BSS_RSP and
Defers all messages. Hence ping stall as tx is disabled and
roam failure not handled.
The fix is to post ADD_BSS_RSP to lim with status code set to TIMEOUT
so that LIM can trigger connection failure to the upper layer and
tear down the link.
Change-Id: I11e19158b9b607c7d9b503305dfc36d519a03077
CRs-Fixed: 610709
|
| | |\ \ \ \ \ |
|
| | |/| | | | |
| | |/ / / /
| | | | | |
| | | | | | |
Release 1.0.0.27 QCACLD WLAN Driver
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change-Id: If7ca96af9553e3b05e8df39120cd41fecb0eaeb6
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Add sanity check to verify the length of the
DFS phyerror event received by the host.
This check will make sure that host is not
flooded with false dfs phyerror events.
CRs-Fixed: 609352
Change-Id: I591e3b67d22d81ae1a50ec624539e88964450c5d
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This fix will resolve static analysis errors reported in wma
module.
Change-Id: I6b203cab47d38bdf0e399050c8e169d5eed93ae1
CRs-Fixed:608831
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Processing Rx data frames upto netcore in softirq context affects
load balancing across available cpu cores. This basically affects
Downlink performance where all the processing right from the interrupt
are done only on CPU0. Do a context switch from tlshim to improve the
performance. This patch introduces a new thread in VOSS for tlshim
data rx.
The thread will be enabled only for SMP kernel. There is an ini
option (gEnableRxThread) to enable disable rx processing in thread.
Thread is enabled by default and this ini is mainly for debugging
purpose, this ini will go away after some time.
Change-Id: I523370fcb6d50b5eee03206c1ce2dafff3b36e8d
CRs-Fixed: 563040
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Add support for vendor specific event and use it to provide
interfering frequency information to userspace. Userspace should
stop P2P-GO, in case P2P-GO is operating on interfering frequency
and restart P2P-GO by avoiding interfering frequency range(s). User
space may decide to continue operation on interfering frequency,
but in such case, there might be impact on performance.
Change-Id: Ib976de3990c0a900cefbe2815e77ed91620d917b
CRs-Fixed: 531882
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Adding modules_install support in Makefile so that external build
entities like bitbake can call the Makefile to install the
modules.
Change-Id: I1956bf09a5f92616cfc12d7c3982644e8765ef9b
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Fixing compiler warnings for uninitialized variables. And also
fixing a warning of function tdlsMsgProcessor exceeding frame
size of 1024.
Change-Id: I756cfa4d68fff75c4446b5281b44f8f9c07de114
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Setting tid value from received packet
to fowarded packet.
CRs-Fixed: 608958
Change-Id: I2161d82612918956cae6f344c257821ca1212316
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This is a new feature support for QXDM logging and RIDL. The
proposed cnss-diag service will handle FW events, debug logs.
Change-Id: I88b69453dc4c890d03ea4d93d97fd0a64f4f5e4b
CRs-fixed: 587068
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change VOS_ASSERT(0) to panic in vos_mem_free() guarded under
MEMORY_DEBUG feature on double free detection.
Change-Id: I9174b885fe643164c9c7b244f1e7a7f47a9cfeb8
CRs-Fixed: 610180
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Increasing demends on 11ac support over 2.4G.
To enable 11ac over 2.4G, we need to set up
-hostapd/supplicant config
-ini value gEnableVhtFor24GHzBand
Change-Id: Id910dc1b303cec9fb306270b7674c8a72a340ccd
CRs-fixed: 608702
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Changes in wmi header files to enable usage of bc_tx_rate
in vdev_start command.
Change-Id: I52fa260301afd32396e40f230dbaddbbc3dc3b57
CRs-Fixed: 609689
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
TX, RX and MC Threads should complete exexution before
the exit routine tries to free the dynamic data these threads
are trying to process. The current implementation does an
interruptible wait for completion of the threads; but it has been
observed that if the process which initiated the driver unload
receives SIGKILL while waiting for threads to complete, the process
will continue without waiting for threads to complete and the threads
will try to access invalidated data after the module has been removed
leading to a kernel panic. To fix this make the wait uninterruptible
and ensure that the driver threads complete execution before the driver
exits. Add similar change for SSR shutdown routine.
CRs-Fixed: 609554
Change-Id: Iba8f5db83fd36ee31381085336b71f50005da7bb
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
There are bunch of header files under WMA/inc which
are unused because the definitions in these files are
taken from WDA/inc files. There are couple of c files
under TXRX which implements some of the TL APIs, since
tl_shim takes care of TL APIs, remoe these files too.
Change-Id: Id79896addd6e008e3ae5d5def3f89407640eb330
CRs-fixed: 592735
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Memory allocated using vos_mem_malloc must needs to be freed by
vos_mem_free otherwise in SLUB debug builds where extra memory
allocated for memStruct meta info is never freed and when memory
actual allocated memory using vos_mem_malloc is freed by calling
adf_os_mem_free SLUB debug enabled kermel complains loudly about
invalid pointer getting freed and in long run of test it leads
to dog bark crash
Change-Id: Idef00457bf58da3cd8c1971221c81d6b6dfc6fd1
CRs-fixed: 609315
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Printing the packet information for tx/rx mgmt, action, EAPOL
and DHCP for debugging the connection issues.
A new ini variable is "gEnableDebugLog" added to
enable/disable these logs. By default this flag is disabled.
Also added an iwpriv command "setDbgLvl" to achieve the same.
Change-Id: Id41af78e036eb6609200d4cd72adff47bf7a25bb
CRs-Fixed: 600142
|
| | |\ \ \ \ \ |
|
| | |/| | | | |
| | |/ / / / |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change-Id: I6a0a322f633c7dc45bd5f597d44b4cdcd1c02be9
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
During SSR, version dump code crashes because of using wrong
device pointer to get the adapter context. Fixed the issue by
getting the adapter context using hardware MAC address.
Change-Id: I54bb5aad952259ad64924c6983f0112ffc48e244
CRs-fixed: 589633
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Sometimes when target assert happens, host driver fails to access
target RAM memory space to copy the dump. Hence add sanity checks
before copy the memory to help identify why it fails.
Also disable dumping AXI registers to avoid potential watchdog bark.
Change-Id: I903d4b2af1721328075fdb8d8fed8686e359ee8d
CRs-Fixed: 604709
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Sometimes it is observed that the number of bytes
of Rx message is very huge which could lead a system
hang. Drop the message indicated with invalid nbytes.
Change-Id: I42497caed98b22f83d83fd51baa129a885f3d1b8
CRs-Fixed: 597089
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
In concurrency scenario, firmware used to send pause and unpause event
to host in order to fairly distribute data flow. The race condition is
that the vdev is stopped when the vdev is already in pause state.
So firmware never sends unpause event for stopped vdev. So the paused
state never be reset and that results in all tx frames is buffered in
txrx itself. Since M2 is the first data frame from next connection,
it never be downloaded to firmware and is causing connection failure.
Change-Id: I58b29c9e50f82eb86aa027140a2e8a408f4c862d
CRs-Fixed: 602021
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This function discards the tx frames queued in a specified vdev.
This applies to HL systems, and also applies to LL systems that
use vdev queues for MCC and/or tx thermal throttling. Also fix bug
of canceling/starting ll_pause.timer without holding a mutex lock.
Change-Id: Ib9833fb6ce42b50a31d83c21c8acfadd1929002a
CRs-Fixed: 602021
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change-Id: Icee29b299cf891874736631f40e558e1f6c1e31a
.
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This reverts commit 85003f4d407e65c8154d2686810e68d894646339
This got a duplicate commit due to missing change-id from git merge
from detached head. This is actually revert of this change which
actually reverted: https://review-android.quicinc.com/#/c/583655/
Change-Id: I3dc379d9f9deb00d8eb1177a3690d11b5a90a5e4
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
When the FW does not send response to the scan command,
the host will be in wait state and will not process any
further commands from the userspace. To avoid this,
timeout is implemented for the scan command on the host
driver. This change is done on the premise that WMA
does not receive any further scan command from the UMAC
until the previous scan command is processed back to the
userspace.
CRs-Fixed: 605502
Depends-on: 598919 588153
Change-Id: I1885fa0f64211b589a2a40f1d858a93235cb7bb2
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Running static source analysis on WMA module generates quite a
few errors/warnings. This fix will help to remove those errors
and warnings
Change-Id: I0db9f17e2635401073a42f4e2de6679eec0a8457
CRs-Fixed: 608831
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
- separate the version dump for wcn/qca device
- introduce the new knob for qwlan_hw for future support,
that way we can differentiate chip version specific info
- example of the output - iwpriv wlan0 version
Host SW:3.2.0.71, FW:2013.12.2, HW:AR6320_REV1_1
- cleanup - remove invalid comment
- add version for 1.1/1.3/2.1
Change-Id: Ifd52d3195a5e56f74322adb1aa6c476889ebee8a
CRs-fixed: 589633
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
For AP auto channel selection, customers should be able to:
1.Only choose 2.4G for scan
2.Only choose 5G for scan
3.Scan both bands
a.Start with 2.4G, if ACS detects a suitable channel,
setup SoftAP in 2.4G itself do not waste time scanning 5G
b.Start with 5G, if ACS detects a suitable channel,
setup SoftAP in 5G itself, do not waste time scanning 2.4G
Two iwpriv commands are added to set and get scan band
preference. When scan preference is set, the preferred
band will be scanned first, if a channel is selected then
the scan is finished, otherwise the next band will be scanned.
Command format:
iwpriv wlan0 scan_band_pref <0|1|2> 0 --- no preference,
1 --- prefer 2.4G band
2 --- prefer 5G band
iwpriv wlan0 get_scan_pref
Change-Id: Ifde9a49f281935bbc141d5a8979f1ca63adabc23
CRs-fixed: 586970
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This feature is to transmit set patterns
at specified intervals by the FW.
Rome firmware should transmit frame(s)
provided by APPS processor periodically.
Linux HDD should export an interface
through which application can set other transmit
parameters to firmware.
Change-Id: I43bf9f47967034d27f9677d4704eb23e388247a1
CRs-fixed: 602128
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
VDEV restart was sent to FW during normal connection
as well as during on channel change. Due to this disconnection
is happening if the new channel from csa is not a dfs channel.
With this fix, following sequence is used VDEV start followed
VDEV UP during normal connection and on channel change VDEV
restart followed by VDEV UP on restart resp handler.
CRs-Fixed: 581476
Depends-on: 588153 597667 598919
Change-Id: I5cccf23fdf7a32a1e0a72b6eb2d05af065c21163
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Changes in WMA to support WMI indication to FW when host is conducting
DHCP operations. By this FW can take action and control BT xSCO actvities.
Change-Id: Iec00a6d5d90c4af973f691ccab63fa82b85cd6c5
CRs-Fixed: 588989
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Changed behavior of opportunistic/periodic roam scan
- Make opportunistic scan rssi same as roam rssi threshold
- Use gEmptyScanRefreshPeriod as its period
- Disable periodic scan when gEmptyScanRefreshPeriod is 0
- When NCHO is enabled use static channel list
- Otherwise use dynamic channel list to enable learning
CRs-Fixed: 600931
Change-Id: I20432709616c4e32e1a198d81d0caea5b0dc1965
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
There are 2 files, Kbuild and Makefile and both have almost
similar content. With this all the Makefile changes are duplicate
between Kbuild and Makefile. Modified Makefile such that Kbuild
will be used for building CLD module.
Change-Id: I0a09919d1b788ac2d397980aa8128349332bcef7
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
In wow mode, host expects firmware to trigger a wake up when certain
discovery packet such as LLMNR, SSDP, mDNS received. So, host adds
discovery pattern as default pattern in STA mode. By mistake, wrong
pattern offet value (38) is configured for discovery packet during
host suspend. Destination mcast ip addr present in discovery packet
starts in offset 30. Due to wrong pattern offset value configuration,
firmware missed to wakeup the host in the reception of discovery packet
This patch sets correct pattern offset value for discovery packet.
Change-Id: Ib5af8c04574c1ae0457511bd285621efc9087225
CRs-Fixed: 607024
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
fixing compiler error for the SAP module by explicit type-casting
for the output of a ternary operator.
CRs-fixed: 582062
Change-Id: Iaba6df91148bc1b2d3f43df0094de0885b1ebc33
|
