During system shutdown, a kernel crash could occur because the r8126
driver's tally counter update routine (`rtl8126_dump_tally_counter`)
was still accessing PCIe address space after the device was already
stopped or powered down.
To prevent this, introduce a new driver flag `R8126_FLAG_SHUTDOWN`
that is set in `rtl8126_shutdown()`. This flag is checked
now before accessing tally counters, ensuring no further
crash during shutdown.
Bug 5620576
Change-Id: I2bb74fc52e712198bfe15b109a2136dc8a2affcd
Signed-off-by: Revanth Kumar Uppala <ruppala@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3482325
Reviewed-by: Ashutosh Jha <ajha@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Ninad Malwade <nmalwade@nvidia.com>
Tested-by: Ninad Malwade <nmalwade@nvidia.com>
Reviewed-by: Shobek Attupurath <sattupurath@nvidia.com>
In the previous release for T264 platforms,the restart lane bringup
logic was controlled through the device tree flag
'nvidia,pcs-rx-eq-sw-ovrd'.
In incremental releases, the flag has been renamed
'nvidia,force-restart-lane-bringup'.
To maintain backward compatibility with older device trees,
the driver now checks for both flags.If either flag is present,
restart lane bringup is executed.
The legacy flag 'nvidia,pcs-rx-eq-sw-ovrd' is deprecated.
Bug 5017313
Change-Id: I24040f508ef776e29a0bb0c1f07d4a74fa4cc8cc
Signed-off-by: Revanth Kumar Uppala <ruppala@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3471233
Reviewed-by: Srinivas Ramachandran <srinivasra@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
A function called hrtimer_setup() was added in Linux v6.13 with the
intent that it would eventually replace hrtimer_init(). In Linux v6.15
the hrtimer_init() function was removed.
Use conftest to call the appropriate API based on what is defined
in the kernel.
Bug 5466808
Change-Id: I1c1c4e81c840a058d8c4c0b1616c87cb8a8a8beb
Signed-off-by: Brad Griffis <bgriffis@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3436079
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Revanth Kumar Uppala <ruppala@nvidia.com>
Problem:
- When a Downstream Port Containment (DPC) software trigger is issued, the LTR_EN bit in the Root Port (RP) is cleared as per PCIe spec.
- However, LTR_EN bit of RTL8126 endpoint (EP) which is being expected to reset is still active and sends Latency Tolerance Reports (LTR) to RP.
- This behavior violates the PCIe spec, as LTR_EN is a non-sticky bit and should be cleared automatically on reset.
- As the RP has LTR disabled but the EP still sends LTR messages, it results in Unsupported Request (UR) errors on the RP.
- These UR errors trigger AER (Advanced Error Reporting) recovery, which includes a Secondary Bus Reset (SBR).
- The SBR causes the PCIe link to go down and come back up, but the EP again starts sending LTRs, leading to a infinite error-recovery loop.
Workaround:
- As a temporary fix, disable the LTR_EN bit in the RTL8126 EP during its probe.
- This prevents the EP from sending LTR messages, thereby avoiding UR errors and breaking the loop of AER recovery.
Impact:
- Disabling LTR prevents the EP from entering the L1.2 low power state.
- However, ASPM is currently not enabled in the system, so this workaround has no impact.
Bug 4869463
Change-Id: Ibf7effaeb0f22e952645ef7bf6a18287264e1463
Signed-off-by: Revanth Kumar Uppala <ruppala@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3420019
Reviewed-by: Manikanta Maddireddy <mmaddireddy@nvidia.com>
Reviewed-by: Ashutosh Jha <ajha@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Previously, this option was disabled because the clang version
used was too old (clang-r370808, clang 10). This option has been
supported since clang-r416183b, clang 12. In order to avoid potential
build errors, this option is re-enabled.
Bug 5289423
Change-Id: I1d0fd5a3dfdff06e95eeca13f85a263922c6ecaf
Signed-off-by: Jian-Min Liu <jianminl@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3371014
Reviewed-by: Ankita Garg <ankitag@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
In Linux v6.15, the timer APIs hrtimer_init() and del_timer() have been
removed. The hrtimer_setup() was added in Linux v6.13 to replace
hrtimer_init() and hrtimer_init() have finally been removed. The
functions del_timer()/del_timer_sync() were renamed to
timer_delete()/timer_delete_sync() in Linux v6.15. Use conftest to
detect these changes and update the drivers as necessary.
JIRA LINQPJ14-47
Change-Id: Id3994900384aad4b91155507cda91e04898ab12c
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3336168
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: Brad Griffis <bgriffis@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: svcacv <svcacv@nvidia.com>
Issue: Below issue observed while running TCP_TX perf test
for longer duration.
nvethernet a808a10000.ethernet: [validate_ctx][1082][type:0x2][loga-0x84d2] dma_txrx: Invalid frame len
nvethernet a808a10000.ethernet mgbe0_0: Tx ring[0] is full
nvethernet a808a10000.ethernet mgbe0_0: Tx ring[0] is full
nvethernet a808910000.ethernet eqos_0: Tx ring[0] is full
This is a regression due to the below change -
https://git-master.nvidia.com/r/c/linux-nv-oot/+/3258684
The issue is that TSO packet is being treated as Non-TSO packet.
pskb_expand_head return code is returned directly where it can
return zero for TSO packet.
Fix:
Removed pskb_expand_head since headers are not getting updated in SW.
HW takes care performing the TSO handling. Also optimized the checks
in TSO path code.
Bug 5175569
Change-Id: I7f48ed32898fec51581bf034b953f0ef7a9913f0
Signed-off-by: Bhadram Varka <vbhadram@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3322354
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Srinivas Ramachandran <srinivasra@nvidia.com>
In Linux v6,15, a 'speed' argument was added to the phy_loopback()
function. Add a conftest test to detect this change and update the
nvethernet driver accordingly. Note that if 'speed' is set to 0 when
calling phy_loopback(), then phy_loopback() behaves the same way as it
did before this argument was added. So by default set speed to 0 for the
nvethernet driver.
JIRA LINQPJ14-47
Change-Id: I55f775e672bfa1a00c9ccbd825c82be1868b0b52
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3330685
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: Revanth Kumar Uppala <ruppala@nvidia.com>
This patch ensures that the Maximum Payload Size (MPS) and Maximum Read Request Size (MRRS) settings of the root port associated with r8168 ethernet endpoint are properly configured after overwritten by kernel when pcie_bus_perf is enabled.
Bug 4607316
Change-Id: I7f7b83f74e4ac2104345bd568d9d2e7c03a1441e
Signed-off-by: Revanth Kumar Uppala <ruppala@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3273562
Reviewed-by: Manikanta Maddireddy <mmaddireddy@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Issue: on resume the mac set speed is not called due
to which the data transmissions are not happening in
alternate suspend resume cycles.
This is because the oldlink status is not set to 1 during
the successful resume, in set_speed_work_func retries
Fix: set oldlink to 1 once setspeed is successful in
function set_speed_work_func, so that next time
ether_adjust_link will call the set speed after resume
Bug 4891730
Change-Id: Iadb426f2cada99387ce180dfa2d0c9ee57e7ccd5
Signed-off-by: Bibhay Ranjan <bibhayr@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3260342
Reviewed-by: Bhadram Varka <vbhadram@nvidia.com>
Reviewed-by: svcacv <svcacv@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Ashutosh Jha <ajha@nvidia.com>
issue: PHY ID allocation is failing during the
stress test of insmod and rmmod of nvethernet
module, since PHYIDs were exhausing during the
module insertion.
fix: deregister fixed link in remove for
freeing up the PHYIDs which were allocated during
probe.
Bug 4754744
Signed-off-by: Narayan Reddy <narayanr@nvidia.com>
Change-Id: I6e71c684d4fcc5bb99a971c7d4d41726998cd8ec
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3245742
Reviewed-by: Bitan Biswas <bbiswas@nvidia.com>
Reviewed-by: Bhadram Varka <vbhadram@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
Reviewed-by: Mohan Thadikamalla <mohant@nvidia.com>
Transmit: Add test_tx_bandwidth_dump_enable support to
transmit packets from the driver which helps measure
line rate for 25G and 10G.
pkt1500, pkt8192, pkt64 packet sizes can be configured.
Receive- Drop Packets at the driver level.
Bug 4742868
Change-Id: I52fe1b73b169dfedc06afea5fb62d7bc21fd9ba1
Signed-off-by: Nagaraj Annaiah <nannaiah@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nv-oot/+/3197198
Reviewed-by: Ashutosh Jha <ajha@nvidia.com>
GVS: buildbot_gerritrpt <buildbot_gerritrpt@nvidia.com>
