summaryrefslogtreecommitdiffstats
path: root/drivers/gpu
Commit message (Collapse)AuthorAge
* gpu: nvgpu: Add PCIe id to device node nameTerje Bergstrom2016-05-05
| | | | | | | | | | | | | Use device name to distinguish between different instances of PCIe devices. JIRA DNVGPU-7 Change-Id: Ice0a9dae41396c8faada1815afd1237907896036 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1140685 Reviewed-by: Sami Kiminki <skiminki@nvidia.com> GVS: Gerrit_Virtual_Submit
* gpu: nvgpu: Add trace and debugfs for sched paramsThomas Fleury2016-05-05
| | | | | | | | | | | | JIRA EVLR-244 JIRA EVLR-318 Change-Id: Ie95f42212dadcf2d0c1737eeb28812afb03b712f Signed-off-by: Thomas Fleury <tfleury@nvidia.com> Reviewed-on: http://git-master/r/1120603 GVS: Gerrit_Virtual_Submit Reviewed-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-by: Ken Adams <kadams@nvidia.com>
* gpu: nvgpu: don't alloc gmmu bufs manuallyKonsta Holtta2016-05-03
| | | | | | | | | | | | | | Use gk20a_gmmu_{alloc,free}() instead of duplicating their code for page tables. The linsim-specific special case is kept as-is. JIRA DNVGPU-23 JIRA DNVGPU-20 Change-Id: I66d772337bad5d081256b13877c4e713ea8b634a Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1139695 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: adapt gk20a_mm_entry for mem_descKonsta Holtta2016-05-03
| | | | | | | | | | | | | | | For upcoming vidmem refactor, replace struct gk20a_mm_entry's contents identical to struct mem_desc, with a struct mem_desc member. This makes it possible to use the page table buffers like the others too. JIRA DNVGPU-23 JIRA DNVGPU-20 Change-Id: I714ee5dcb33c27aaee932e8e3ac367e84610b102 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1139694 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: cancel clean up before update_fn_workDeepak Nibade2016-05-03
| | | | | | | | | | | | | | | | | | | | | | | In gk20a_channel_suspend(), we first cancel worker thread update_fn_work and then cancel job clean up worker But it is possible that we re-schedule update_fn_work from job cleanup worker after it was cancelled And in that case worker update_fn_work might run after we poweroff GPU Fix this by first cancelling job clean up worker and then update_fn_work worker Bug 200187905 Change-Id: Ia192c515702f14becf60d92c6471d8c0e892551e Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120426 (cherry picked from commit b132b6c5aa9ab88c6733f36906f1b874114ec72d) Reviewed-on: http://git-master/r/1134888 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: remove wait in shutdownDeepak Nibade2016-05-03
| | | | | | | | | | | | | | | | | | In gk20a_pm_shutdown(), we currently wait for 2s for all channels to finish their work But we already cancel all the nvgpu workers, freeze user processes during shutdown So the waiting should not be required, and hence remove it Bug 200166139 Change-Id: I0012f1b3c0f4f676958d083f8c60a001f7015fb0 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1121918 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: return from worker if gpu is not upDeepak Nibade2016-05-03
| | | | | | | | | | | | | | | | | | | | During GPU shutdown path, it is possible that we shut down the GPU while worker thread is still running gk20a_channel_update() Hence before accessing gp_put/get, check if GPU is up or not Bug 200166139 Change-Id: Iba3ec173041a84527c4700a93f20564a842cfb01 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/935193 (cherry picked from commit c81ea5fe383c44e872754b363968af57d84225ac) Reviewed-on: http://git-master/r/1121917 Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: fix resource leaks with gr_gk20a_elpg_protected_callDeepak Nibade2016-05-02
| | | | | | | | | | | | | | | | | | | | | | | | | In gr_gk20a_elpg_protected_call(), we return with error value if we fail to disable elpg But since this is a #define'd function, we end up returning from function which is using gr_gk20a_elpg_protected_call() So in some cases it is possible that parent function does not free up resources due to return statement in gr_gk20a_elpg_protected_call() Fix this by removing return statement, and execute rest of the code if there is no error Coverity id : 31980 Bug 200192125 Change-Id: Ic003b160b76820cdf9355f44658c23bfb2f3815f Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1133404 GVS: Gerrit_Virtual_Submit Reviewed-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Clear FIFO intr before processing finished jobsAlex Waterman2016-04-29
| | | | | | | | | | | | | | | Clear the FIFO interrupt before prcessing the job list after receiving a nonstalling interrupt. This prevents a race in which some non-stalling interrupts after a semaphore incr can get lost. Bug 1732449 JIRA DNVGPU-12 Change-Id: I03df56b2ebca4ed8a0aeb26dd5480c91ffb42d8b Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: http://git-master/r/1133791 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Do not generate any ctag info unless enabledAlex Waterman2016-04-29
| | | | | | | | | | | | | | Do not put any ctag data in the PTEs unless compression is actually enabled for the mapping. Bug 1732449 JIRA DNVGPU-12 Change-Id: I2abfbf9d1282af24541f8199bd9fbf2133c12899 Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: http://git-master/r/1133790 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Wake up the correct workqueueAlex Waterman2016-04-29
| | | | | | | | | | | | | | | | | Wake up the correct workqueue during the nonstalling interrupt handler. Previously the stalling workqueue was woken up which lead to any process waiting on the nonstalling workqueue hanging indefinitely. Bug 1732449 JIRA DNVGPU-12 Change-Id: I8744ceddd7957bbaee0b8203f9a3aaf8ad3792fc Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: http://git-master/r/1133788 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Flush FB before checking semaphoresAlex Waterman2016-04-29
| | | | | | | | | | | | | | | | | Before checking semaphore values to determine if jobs have been completed flush the FB. If this is not done, despite the sempahore memory being mapped as volatile in the GMMU, outstanding writes can still be pending. Bug 1732449 JIRA DNVGPU-12 Change-Id: I67b596cd23a5465af05d6d173641a579cb7f168c Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: http://git-master/r/1133787 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Schedule channel update when jobs actually finishAlex Waterman2016-04-29
| | | | | | | | | | | | | | | Do not schedule channel update call backs unless a job is actually finished. This saves a lot of call backs to the CDE code that don't do anything when semaphores are enabled. Bug 1732449 JIRA DNVGPU-12 Change-Id: I2f9a78498b08ebca44ee6a5171931a07721767f1 Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: http://git-master/r/1133786 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Add gk20a_gmmu_fixed_map() functionAlex Waterman2016-04-29
| | | | | | | | | | | | | | | | Add a function to allow the kernel to do fixed mappings. Necessary for the semaphore functionality since there needs to be a common address in each VM for the semaphores. Bug 1732449 JIRA DNVGPU-12 Change-Id: I2b451db2d3cb3c003d951f7b0ffc87f6c91db7dc Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: http://git-master/r/1133789 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Allocate channel table with vmallocTerje Bergstrom2016-04-28
| | | | | | | | | | | | | | | | | Channel table can be bigger than one page, so allocate it with vmalloc. Also add a free for tsg table, which did not exist before, and remove per-channel remove_channel callback which was never used. JIRA DNVGPU-50 Change-Id: I3ee84b65d94881df52bf0618bf4c5f2e85758223 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1129244 Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Ken Adams <kadams@nvidia.com> GVS: Gerrit_Virtual_Submit
* gpu: nvgpu: add TSG timeslice supportAingara Paramakuru2016-04-27
| | | | | | | | | | | | | | | Add support for changing a TSG's timeslice, within reasonable limits imposed by the kernel driver. JIRA VFND-1494 Bug 1749744 Change-Id: Ifca1b63a00da7a5872483bb56692da70a5f18bdf Signed-off-by: Aingara Paramakuru <aparamakuru@nvidia.com> Reviewed-on: http://git-master/r/1129837 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Fix floorsweeping for multi-GPC GPUTerje Bergstrom2016-04-27
| | | | | | | | | | | | | There were multiple bugs in dealing with a GPU with more than one GPC. * Beta CB size was set to wrong PPC * TPC mask did not shift fields correctly * PD skip table used || instead of | operator Change-Id: I849e2331a943586df16996fe573da2a0ac4cce19 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1132109
* gpu: nvgpu: Add PCIe device supportTerje Bergstrom2016-04-27
| | | | | | | | | | Add support for probing PCIe graphics cards. JIRA DNVGPU-7 Change-Id: Iad3d31a1dc0ca6575d8a9916857022cac9181948 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1127684
* gpu: nvgpu: Idle GR before calling PMU ZBC saveTerje Bergstrom2016-04-26
| | | | | | | | | | | | | | | On gk20a when PMU is updating ZBC colors it is reading them from L2. But L2 has one port, and ZBC reads can race with other transactions. Idle graphics before sending PMU the ZBC_UPDATE request. Also makes pmu_save_zbc a HAL, because PMU ucode has changes to bypass this problem on some chips. Bug 1746047 Change-Id: Id8fcd6850af7ef1d8f0a6aafa0fe6b4f88b5f2d9 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1129017
* gpu: nvgpu: Program NISO sysmem flush addrTerje Bergstrom2016-04-25
| | | | | | | | | | | | Program sysmem flush address to prevent random accesses of address 0. Change-Id: I886170395f036805f02e0bce7ecd3c8c46b921df Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1129216 GVS: Gerrit_Virtual_Submit Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-by: Konsta Holtta <kholtta@nvidia.com>
* gpu: nvgpu: do not include hw_proj_*.hDeepak Nibade2016-04-23
| | | | | | | | | | | | | | | | | | hw_proj_gk20a.h and hw_proj_gm20b.h should not be included, hence remove the includes and APIs used from the header Use nvgpu_get_litter_value() API to replace use of header Bug 200156699 Change-Id: I5e88f71657682dd94ac7f0a45f940b70cf8222e7 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1129611 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: set soc memory aperture typeSeshendra Gadagottu2016-04-23
| | | | | | | | | | | | | | | Set platform data for soc memory aperture type, whether soc memory aperture seen as sysmem or vidmem. For gk20a/gm20b, soc memory aperture seen as vidmem. Bug 1749338 Change-Id: I407562ca484c1a4bae1bee12089d2b19f378ca53 Signed-off-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-on: http://git-master/r/1129167 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Ken Adams <kadams@nvidia.com>
* gpu: nvgpu: fix sparse warningsDeepak Nibade2016-04-20
| | | | | | | | | | | | | | | | | | | | | Fix below sparse warnings : drivers/gpu/nvgpu/gk20a/gk20a.c:764:5: warning: symbol 'gk20a_pm_finalize_poweron' was not declared. Should it be static? drivers/gpu/nvgpu/gk20a/channel_gk20a.c:2504:14: warning: symbol 'gk20a_event_id_poll' was not declared. Should it be static? drivers/gpu/nvgpu/gk20a/channel_gk20a.c:2538:5: warning: symbol 'gk20a_event_id_release' was not declared. Should it be static? Bug 200067946 Bug 200088648 Change-Id: I5c23e7ee09c1a18fe2eeff12f80a3c2bf73120ef Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1128060 Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Sri Krishna Chowdary <schowdary@nvidia.com>
* gpu: nvgpu: assign own platform data for t19xSeshendra Gadagottu2016-04-19
| | | | | | | | | | | Bug 1735765 Change-Id: I6adf9cbe8ba636d5e05e2aa3ac46f7f20b1de7ed Signed-off-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-on: http://git-master/r/1128303 Reviewed-by: Ken Adams <kadams@nvidia.com> Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit
* gpu: nvgpu: make jobs_lock more fine grainedDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | | | | | While processing all the jobs in gk20a_channel_clean_up_jobs(), We currently acquire jobs_lock, traverse the list, clean up the jobs, and then release the lock But in this case we might hold the lock for too long blocking the submit path Hence make jobs_lock more fine grained by restricting it for list accesses only Bug 200187553 Change-Id: If82af8ff386f7bc29061cfd57fdda7df62f11c17 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120412 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: remove submit lockDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | | | Remove submit lock since we have moved to use more fine-grained locks Remove API check_gp_put() since we cannot call it in submit path due to latencies and we cannot call it in gk20a_channel_clean_up_jobs() anymore since it will fail there without the lock Bug 200187553 Change-Id: I05b9fa95c9009000e13232d8fa567336eeee11c6 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120411 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: implement sync refcountingDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | We currently free sync when we find job list empty If aggressive_sync is set to true, we try to free sync during channel unbind() call But we rarely free sync from channel_unbind() call since freeing it when job list is empty is aggressive enough Hence remove sync free code from channel_unbind() Implement refcounting for sync: - get a refcount while submitting a job (and allocate sync if it is not allocated already) - put a refcount while freeing the job - if refcount==0 and if aggressive_sync_destroy is set, free the sync - if aggressive_sync_destroy is not set, we will free the sync during channel close time Bug 200187553 Change-Id: I74e24adb15dc26a375ebca1fdd017b3ad6d57b61 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120410 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: add lock for fencesDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | All pre/post fence accesses in last_submit are currently protected by submit lock In order to remove the submit lock, move all fence accesses under own lock i.e. fence_lock Bug 200187553 Change-Id: I0132d1933dc92db8c5ed8c9311e49a030aa2d38c Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120409 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: add accessors for global_esr values and sm_dbgr_controlDeepak Nibade2016-04-19
| | | | | | | | | | | | | Add gk20a/gm20b accessors for various global_esr values and for sm_dbgr_control modes Bug 200156699 Change-Id: If7fd8cd7567f8bcd1f645facf9553bdc0a153526 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120333 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: IOCTL to suspend/resume contextDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Add below IOCTL to suspend/resume a context NVGPU_DBG_GPU_IOCTL_SUSPEND_RESUME_CONTEXTS: Suspend sequence : - disable ctxsw - loop through list of channels - if channel is ctx resident, suspend all SMs - otherwise, disable channel/TSG - enable ctxsw Resume sequence : - disable ctxsw - loop through list of channels - if channel is ctx resident, resume all SMs - otherwise, enable channel/TSG - enable ctxsw Bug 200156699 Change-Id: Iacf1bf7877b67ddf87cc6891c37c758a4644b014 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120332 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: support binding multiple channels to a debug sessionDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We currently bind only one channel to a debug session But some use cases might need multiple channels bound to same debug session Add this support by adding a list of channels to debug session. List structure is implemented as struct dbg_session_channel_data List node dbg_s_list_node is currently defined in struct dbg_session_gk20a. But this is inefficient when we need to add debug session to multiple channels Hence add new reference structure dbg_session_data to store dbg_session pointer and list entry For each NVGPU_DBG_GPU_IOCTL_BIND_CHANNEL call, create two reference structure dbg_session_channel_data for channel and dbg_session_data for debug session and bind them together Define API nvgpu_dbg_gpu_get_session_channel() which will get first channel in the list of debug session Use this API wherever we refer to channel bound to debug session Remove dbg_sessions define in struct gk20a since it is not being used anywhere Add new API NVGPU_DBG_GPU_IOCTL_UNBIND_CHANNEL to support unbinding of channel from debug sesssion Bug 200156699 Change-Id: I3bfa6f9cd5b90e7254a75c7e64ac893739776b7f Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120331 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu; nvgpu: IOCTL to write/clear SM error statesDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | Add below IOCTLs to write/clear SM error states NVGPU_DBG_GPU_IOCTL_CLEAR_SINGLE_SM_ERROR_STATE NVGPU_DBG_GPU_IOCTL_WRITE_SINGLE_SM_ERROR_STATE Bug 200156699 Change-Id: I89e3ec51c33b8e131a67d28807d5acf57b3a48fd Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120330 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: support storing/reading single SM error stateDeepak Nibade2016-04-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Add support to store error state of single SM before preprocessing SM exception Error state is stored as : struct nvgpu_dbg_gpu_sm_error_state_record { u32 hww_global_esr; u32 hww_warp_esr; u64 hww_warp_esr_pc; u32 hww_global_esr_report_mask; u32 hww_warp_esr_report_mask; } Note that we can safely append new fields to above structure in the future if required Also, add IOCTL NVGPU_DBG_GPU_IOCTL_READ_SINGLE_SM_ERROR_STATE to support reading SM's error state by user space Bug 200156699 Change-Id: I9a62cb01e8a35c720b52d5d202986347706c7308 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1120329 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: fix interface name for ctxsw devnodeThomas Fleury2016-04-18
| | | | | | | | | | | | gk20a_create_device now takes interface name as a parameter. This patche adds it for ctxsw devnode. Change-Id: Ic076159829574c6015056fb950f625a2233c015f Signed-off-by: Thomas Fleury <tfleury@nvidia.com> Reviewed-on: http://git-master/r/1123721 Reviewed-by: Richard Zhao <rizhao@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: add TSG interleave supportAingara Paramakuru2016-04-16
| | | | | | | | | | | | | | Add support for changing a TSG's runlist interleave level. JIRA VFND-1497 Bug 1749744 Change-Id: I3cf3ebc2334f83b1bfb6b3230fae2ca73c75c239 Signed-off-by: Aingara Paramakuru <aparamakuru@nvidia.com> Reviewed-on: http://git-master/r/1122677 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Implement NVGPU_GPU_IOCTL_GET_GPU_TIMESami Kiminki2016-04-15
| | | | | | | | | | | | Implement NVGPU_GPU_IOCTL_GET_GPU_TIME for reading the GPU time. Bug 1395833 Change-Id: I7ddc7c28ff0c9a336cc0dcd820b15fb0fea714d0 Signed-off-by: Sami Kiminki <skiminki@nvidia.com> Reviewed-on: http://git-master/r/1125630 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: gm20b: update pmu verVijayakumar2016-04-15
| | | | | | | | | | | | | | | | bug 1736601 PMU fixes ELPG residency calculation Change-Id: I3fd802b582afe92627da218b7258ac8d307b1d30 Signed-off-by: Vijayakumar <vsubbu@nvidia.com> Reviewed-on: http://git-master/r/1118262 (cherry picked from commit bda3a675b5463f021869015bbf628553ed422325) Reviewed-on: http://git-master/r/1127491 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: WARN_ON if NULL params in pmu_cmd_postGagan Grover2016-04-15
| | | | | | | | | | | | | | | | | | | | | Don't need to BUG_ON. We can simply print kernel error message along with call stack and return. Bug 200182457 Change-Id: I06693f88372dfb5dd0dd2fae7630540594f302ba Signed-off-by: Gagan Grover <ggrover@nvidia.com> Reviewed-on: http://git-master/r/1115992 (cherry picked from commit f5cc88880e147f90d2cf9980db4322ee5fd613a6) Reviewed-on: http://git-master/r/1117166 (cherry picked from commit ab0318b4a726886cd3e5ce97cf1a3bcaabab6171) Reviewed-on: http://git-master/r/1119038 (cherry picked from commit 9b977073d11d7a1caa72994f750293b0e5076cef) Reviewed-on: http://git-master/r/1127490 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Wait for BAR1 bindTerje Bergstrom2016-04-15
| | | | | | | | | Wait for BAR1 bind to complete before continuing. The register to wait exists Maxwell onwards. Change-Id: Ie3736033fdb748c5da8d7a6085ad6d63acaf41f5 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1123941
* gpu: nvgpu: Use sysmem aperture for SoC memoryTerje Bergstrom2016-04-15
| | | | | | | | | In Tegra GPU, SoC memory has to be accessed as vidmem. In discrete GPU, it has to be accessed as sysmem. Change-Id: I4efe71b54a9a32f0bf1f02ec4016ed74405a14c5 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1120468
* gpu: nvgpu: Add litter values HALTerje Bergstrom2016-04-15
| | | | | | | | | Move per-chip constants to be returned by a chip specific function. Implement get_litter_value() for each chip. Change-Id: I2a2730fce14010924d2507f6fa15cc2ea0795113 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: http://git-master/r/1121383
* gpu: nvgpu: pass device pointer as argument to nvhostDeepak Nibade2016-04-15
| | | | | | | | | | | | | | Pass device pointer as an argument to nvhost_register_dump_device() This is needed so that nvhost can call the callback with device pointer that we register Bug 200188753 Change-Id: I4e3ff7e99378b9d7ec25b88a7af2c89998496b4c Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1126791 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: support kernel-3.10 versionDeepak Nibade2016-04-15
| | | | | | | | | | | | | | | | | | | | | | Make necessary changes to support nvgpu on kernel-3.10 This includes below changes - PROBE_PREFER_ASYNCHRONOUS is defined only for K3.10 - Fence handling and struct sync_fence is different between K3.10 and K3.18 - variable status in struct sync_fence is atomic on K3.18 whereas it is int on K3.10 - if SOC == T132, set soc_name = "tegra13x" - ioremap_cache() is not defined on K3.10 ARM versions, hence use ioremap_cached() Bug 200188753 Change-Id: I18d77eb1404e15054e8510d67c9a61c0f1883e2b Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1121092 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: use get_unused_fd_flags()Deepak Nibade2016-04-15
| | | | | | | | | | | | | | | get_unused_fd() is deprecated on kernel-4.4, but get_unused_fd_flags() is supported on both kernel-4.4 and previous versions hence remove use of get_unused_fd() and use get_unused_fd_flags() Change-Id: I132aa67d2bc23a698848ac51d2d176d7d33e1695 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1126847 GVS: Gerrit_Virtual_Submit Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>
* gpu: nvgpu: make tpc_fs_mask work on production boardRichard Zhao2016-04-14
| | | | | | | | | | | | | | | On production fused boards, it uses gr_fe_tpc_fs_r() to mask TPCs, rather than fues. Bug 1734150 Change-Id: I7b4eb428f1ad0cf841a57214e0c8c1e8f17b2c5a Signed-off-by: Richard Zhao <rizhao@nvidia.com> Reviewed-on: http://git-master/r/1111630 (cherry picked from commit 869ea54967812e03d9f1e69775ca56fd6459216c) Reviewed-on: http://git-master/r/1122121 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: vgpu: support reduced BAR1 rangeAingara Paramakuru2016-04-14
| | | | | | | | | | | | | | | Going forward, only the guest-accessible BAR1 range will be contained in the DT. Full-range BAR1 support is maintained for backwards-compatibility. JIRA VFND-1373 Change-Id: I13cb7e3b76397f76788838c6debcdafa43a8ea25 Signed-off-by: Aingara Paramakuru <aparamakuru@nvidia.com> Reviewed-on: http://git-master/r/1126380 Reviewed-by: Richard Zhao <rizhao@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Vladislav Buzov <vbuzov@nvidia.com>
* gpu: nvgpu: clean up ctx_vars properlyKonsta Holtta2016-04-14
| | | | | | | | | | | | | | Set ctx_vars.valid to false when removing support. Otherwise a re-poweron sequence could crash when the flag wouldn't match the real state of the driver. Also free all allocated regs instead of leaking some of them. Change-Id: I3fc4fa759d839bc435e53cbd942fa5d39efe7f57 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1126138 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: free global ctx bufs only if allocatedKonsta Holtta2016-04-14
| | | | | | | | | | | | | | | | | Try to free only allocated buffers in gr_gk20a_free_global_ctx_buffers(), otherwise the destroy function pointer would be NULL and crash for nonallocated buffers. This can happen when init fails for some of the buffers. Additionally, make the pointer NULL when a buffer is destroyed, to signify this state. Also refactor the function upwards and call it from gr_gk20a_alloc_global_ctx_buffers() to reduce code duplication. Change-Id: I6e74795014f5e315b5f8342f544ddfccc0d02b71 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1126026 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: move cs_lock init to happen alwaysKonsta Holtta2016-04-14
| | | | | | | | | | | | The cs_lock cyclestats mutex is unconditionally taken when removing cs support, but it wouldn't be initialized if some part of gr init would fail before it. Move it up to happen first, before other inits. Change-Id: Ia5d7a888c29dc99728630a07698b1ed25af960c2 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1126004 Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: fix compilation error with CONFIG_PM disabledDeepak Nibade2016-04-14
| | | | | | | | | | | gk20a_gpu_is_virtual() needs to pass struct device *dev and not pdev which is undefined Change-Id: I8835bb1175efa693b468588e91aaef9e5531d0bc Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: http://git-master/r/1125439 GVS: Gerrit_Virtual_Submit Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>