summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/nvgpu/gk20a/channel_gk20a.h
Commit message (Collapse)AuthorAge
* gpu: nvgpu: move channel header to commonKonsta Holtta2018-09-05
| | | | | | | | | | | | | channel_gk20a is clear from chip specifics and from most dependencies, so move it under the common directory. Jira NVGPU-967 Change-Id: I41f2160b96d4ec84064288ecc22bb360e82352df Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1810578 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: move priv cmd definitions to channelKonsta Holtta2018-08-31
| | | | | | | | | | | | | | struct priv_cmd_queue and struct priv_cmd_entry are related to the list of jobs in a channel, so move their definitions from the mm header to the channel header. Jira NVGPU-967 Change-Id: Ib0cf3fd52be463e720165a47e56b14724273473e Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1807371 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: move gpfifo_desc to channel headerKonsta Holtta2018-08-31
| | | | | | | | | | | | | struct gpfifo_desc belongs to channel; reduce dependencies to mm by moving it there. Jira NVGPU-967 Change-Id: I3c6ac3d4e47fc0c843d407ee0e1100a009763d08 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1807398 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: remove ctx header desc typeKonsta Holtta2018-08-31
| | | | | | | | | | | | | | The graphics subctx header object is nothing but memory. Drop the dependency to gr header file in the channel header file and substitute struct nvgpu_mem for struct ctx_header_desc. Jira NVGPU-967 Change-Id: Ic3976391016c42d2ada4aac3e0851a1222244ce9 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1807370 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: use own usermode gpfifo objectKonsta Holtta2018-08-30
| | | | | | | | | | | | | | | | | | | Add usermode_gpfifo member to channel to store the usermode-owned buffer data. This makes it more explicit about which (user or kernel) memory is accessed. Jira NVGPU-967 Change-Id: I98e4f4568f04aa31e00e497794fadd4997ba8aab Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1807503 Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: clean up channel header includesKonsta Holtta2018-08-24
| | | | | | | | | | | | | | | | Remove a few unnecessary includes from channel_gk20a.h and add them to c files where needed. Jira NVGPU-967 Change-Id: Ic38132c776a56b6966424806faab7871575b6c10 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1804609 Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Reduce structure padding wasteNicolin Chen2018-08-22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The gk20a_init_fifo_setup_sw_common() function allocates memory of schannel_gk20a and tsg_gk20a tructures for all 512 channels: Size Caller Module Pages Type 749568 __nvgpu_vzalloc+0x28/0x78 [nvgpu] pages=182 vmalloc 602112 __nvgpu_vzalloc+0x28/0x78 [nvgpu] pages=146 vmalloc This change just simply reorgnizes the member defines in those two structures to reduce padding waste. After this change: Size Caller Module Pages Type 733184 __nvgpu_vzalloc+0x28/0x78 [nvgpu] pages=178 vmalloc 585728 __nvgpu_vzalloc+0x28/0x78 [nvgpu] pages=142 vmalloc In summary, it saves 8 pages in 32KB memory. Bug 2327574 Bug 2284925 Change-Id: I06693e0fef516a145b48dd3a05d756c0feaf3ba5 Signed-off-by: Nicolin Chen <nicolinc@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1803358 Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com> Reviewed-by: svccoveritychecker <svccoveritychecker@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Alex Waterman <alexw@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Conditional enable for replayable faultTerje Bergstrom2018-07-09
| | | | | | | | | | | | | | | | Enable replayable fault only for contexts where they are requested. This required moving the code to initialize subcontexts to happen later. Fix signedness issues in definition of flags. JIRA NVGPU-714 Change-Id: I472004e13b1ea46c1bd202f9b12d2ce221b756f9 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1773262 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: move submit code to commonKonsta Holtta2018-06-27
| | | | | | | | | | | | | | | | To finish OS unification of the submit path, move the gk20a_submit_channel_gpfifo* functions to a file that's accessible also outside Linux code. Also change the prefix of the submit functions from gk20a_ to nvgpu_. Jira NVGPU-705 Change-Id: I8ca355d1eb69771fb016c7a21fc7f102ca7967d7 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1760421 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: recover on first interrupt reported for mmu nackAparna Das2018-06-15
| | | | | | | | | | | | | | | | | | | In case of mmu nack error interrupt is received twice through SM reported mmu nack interrupt and mmu fault in undertermined order. Recover on the first received interrupt to avoid semaphore release and skip doing a second recovery. Also fix NULL pointer dereference in function gv11b_fifo_reset_pbdma_and_eng_faulted when channel reference is invalid in teardown path. Bug 200382235 Change-Id: I361a5725d7b6355ebf02b2870727f647fbd7a37e Signed-off-by: Aparna Das <aparnad@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1739804 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: add NVGPU_IOCTL_CHANNEL_RESCHEDULE_RUNLISTDavid Li2018-05-18
| | | | | | | | | | | | | | | | | | | | | | | Add NVGPU_IOCTL_CHANNEL_RESCHEDULE_RUNLIST ioctl to reschedule runlist, and optionally check host and FECS status to preempt pending load of context not belonging to the calling channel on GR engine during context switch. This should be called immediately after a submit to decrease worst case submit to start latency for high interleave channel. There is less than 0.002% chance that the ioctl blocks up to couple miliseconds due to race condition of FECS status changing while being read. For GV11B it will always preempt pending load of unwanted context since there is no chance that ioctl blocks due to race condition. Also fix bug with host reschedule for multiple runlists which needs to write both runlist registers. Bug 1987640 Bug 1924808 Change-Id: I0b7e2f91bd18b0b20928e5a3311b9426b1bf1848 Signed-off-by: David Li <davli@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1549050 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: sync_framework cleanupsDebarshi Dutta2018-04-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch deals with cleanups meant to make things simpler for the upcoming os abstraction patches for the sync framework. This patch causes some substantial changes which are listed out as follows. 1) sync_timeline is moved out of gk20a_fence into struct nvgpu_channel_linux. New function pointers are created to facilitate os independent methods for enabling/disabling timeline and are now named as os_fence_framework. These function pointers are located in the struct os_channel under struct gk20a. 2) construction of the channel_sync require nvgpu_finalize_poweron_linux() to be invoked before invocations to nvgpu_init_mm_ce_context(). Hence, these methods are now moved away from gk20a_finalize_poweron() and invoked after nvgpu_finalize_poweron_linux(). 3) sync_fence creation is now delinked from fence construction and move to the channel_sync_gk20a's channel_incr methods. These sync_fences are mainly associated with post_fences. 4) In case userspace requires the sync_fences to be constructed, we try to obtain an fd before the gk20a_channel_submit_gpfifo() instead of trying to do that later. This is used to avoid potential after effects of duplicate work submission due to failure to obtain an unused fd. JIRA NVGPU-66 Change-Id: I42a3e4e2e692a113b1b36d2b48ab107ae4444dfa Signed-off-by: Debarshi Dutta <ddutta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1678400 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: pass pid/tid from os specific code to common codeRichard Zhao2018-04-16
| | | | | | | | | | | | | | | | | | linux driver runs in user's process but qnx driver has dedicate driver process, so they have different way to get user pid. nvgpu common code expect calls from os specific code pass pid/tid. ce/cde open channel for internal use, we use driver pid. Jira VQRM-3534 Change-Id: I892372ac5f1dc4d25f9928d16992bcc659d12a56 Signed-off-by: Richard Zhao <rizhao@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1694145 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Add conversion function for uapi submit gpfifo flagsSourab Gupta2018-04-11
| | | | | | | | | | | | | | | | | | The submit gpfifo flags are splattered everywhere inside the nvgpu code. Though the usage is inside nvgpu Linux code only, still it needs to be gotten rid of and replaced with the defines present in common code. VQRM-3465 Change-Id: I901b33565b01fa3e1f9ba6698a323c16547a8d3e Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1691979 Reviewed-by: Konsta Holtta <kholtta@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: remove usage of nvgpu_gpfifoSourab Gupta2018-04-11
| | | | | | | | | | | | | | | | | | | Remove the usage of nvgpu_gpfifo splattered across nvgpu, and replace with a struct defined in common code. The usage is still inside Linux, but this helps the subsequent unification efforts, e.g. to unify the submit path. VQRM-3465 Change-Id: I9e5ac697a0c7f85239ddba319085c09481d20d6b Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1691978 Reviewed-by: Konsta Holtta <kholtta@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: remove usage of nvgpu_fenceSourab Gupta2018-04-11
| | | | | | | | | | | | | | | | | | | Remove the usage of nvgpu_fence splattered across nvgpu, and replace with a struct defined in common code. The usage is still inside Linux, but this helps the subsequent unification efforts, e.g. to unify the submit path. VQRM-3465 Change-Id: Ic3737450123dfc5e1c40ca5b6b8d8f6b3070aa0d Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1691977 Reviewed-by: Konsta Holtta <kholtta@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: add usermode submission interface HALSourab Gupta2018-04-05
| | | | | | | | | | | | | | | | | | | | The patch adds the HAL interfaces for handling the usermode submission, particularly allocating channel specific usermode userd. These interfaces are currently implemented only on QNX, and are created accordingly. As and when linux adds the usermode submission support, we can revisit them if any further changes are needed. Change-Id: I790e0ebdfaedcdc5f6bb624652b1af4549b7b062 Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1683392 Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: pass alloc_gpfifo args to gk20a_channel_alloc_gpfifoSourab Gupta2018-04-05
| | | | | | | | | | | | | | | | The patch defines 'struct nvgpu_gpfifo_args' to be filled by alloc_gpfifo(_ex) ioctls and passed to the gk20a_channel_alloc_gpfifo function. This is required as a prep towards having the usermode submission support in the core channel core. Change-Id: I72acc00cc5558dd3623604da7d716bf849f0152c Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1683391 GVS: Gerrit_Virtual_Submit Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: make fifo/ch functions called by RM Server globalSourab Gupta2018-03-29
| | | | | | | | | | | | | | | | | | The patch declares globally few channel/fifo HAL functions required for QNX code compilation (as they are being referred elsewhere in QNX code). This is required as a part of bringing in the nvgpu Channel/FIFO HAL into QNX. Jira VQRM-3058 Change-Id: Ia176535b64de981d2f7ddb20f62015a0da74fd2a Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1662411 GVS: Gerrit_Virtual_Submit Tested-by: Richard Zhao <rizhao@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: delete unused job->pre_fenceKonsta Holtta2018-03-19
| | | | | | | | | | | | | | | | | | | The pre_fence member in channel_gk20a_job is no longer used for anything. Delete it. Only the post fence needs to be tracked. Jira NVGPU-527 Jira NVGPU-528 Bug 200390539 Change-Id: Ia1a556728dabf9a8e305ed76020ac1aa0b4d6b88 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1676735 Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: support per-channel wdt timeoutsKonsta Holtta2018-03-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Replace the padding in nvgpu_channel_wdt_args with a timeout value in milliseconds, and add NVGPU_IOCTL_CHANNEL_WDT_FLAG_SET_TIMEOUT to signify the existence of this new field. When the new flag is included in the value of wdt_status, the field is used to set a per-channel timeout to override the per-GPU default. Add NVGPU_IOCTL_CHANNEL_WDT_FLAG_DISABLE_DUMP to disable the long debug dump when a timed out channel gets recovered by the watchdog. Printing the dump to serial console takes easily several seconds. (Note that there is NVGPU_TIMEOUT_FLAG_DISABLE_DUMP about ctxsw timeout separately for NVGPU_IOCTL_CHANNEL_SET_TIMEOUT_EX as well.) The behaviour of NVGPU_IOCTL_CHANNEL_WDT is changed so that either NVGPU_IOCTL_CHANNEL_ENABLE_WDT or NVGPU_IOCTL_CHANNEL_DISABLE_WDT has to be set. The old behaviour was that other values were silently ignored. The usage of the global default debugfs-controlled ch_wdt_timeout_ms is changed so that its value takes effect only for newly opened channels instead of in realtime. Also, zero value no longer means that the watchdog is disabled; there is a separate flag for that after all. gk20a_fifo_recover_tsg used to ignore the value of "verbose" when no engines were found. Correct this. Bug 1982826 Bug 1985845 Jira NVGPU-73 Change-Id: Iea6213a646a66cb7c631ed7d7c91d8c2ba8a92a4 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1510898 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: allocate separate client managed syncpoint for UserDeepak Nibade2018-03-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We right now allocate a nvgpu managed syncpoint in c->sync and share that with user space But to avoid conflicts between user space and kernel space increments allocate a separate "client managed" syncpoint for User space in c->user_sync Add new API nvgpu_nvhost_get_syncpt_client_managed() to request a client managed syncpoint from nvhost. Note that nvhost/nvgpu do not keep track of MAX/threshold value of this syncpoint Update gk20a_channel_syncpt_create() to receive a flag to indicate whether a User space syncpoint is required or not Unset NVGPU_SUPPORT_USER_SYNCPOINT for gp10b since we don't want to allocate double syncpoints per channel on that platform For gv11b, once we move to use user space submits, support for c->sync will be dropped so we keep using only one syncpoint per channel Bug 200326065 Jira NVGPU-179 Change-Id: I78d94de4276db1c897ea2a4fe4c2db8b2a179722 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1665828 Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Fold T19x code back to main code pathsTerje Bergstrom2018-01-23
| | | | | | | | | | | | | Lots of code paths were split to T19x specific code paths and structs due to split repository. Now that repositories are merged, fold all of them back to main code paths and structs and remove the T19x specific Kconfig flag. Change-Id: Id0d17a5f0610fc0b49f51ab6664e716dc8b222b6 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1640606 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Make graphics context property of TSGTerje Bergstrom2018-01-17
| | | | | | | | | | | | | | | | | | | | Move graphics context ownership to TSG instead of channel. Combine channel_ctx_gk20a and gr_ctx_desc to one structure, because the split between them was arbitrary. Move context header to be property of channel. Bug 1842197 Change-Id: I410e3262f80b318d8528bcbec270b63a2d8d2ff9 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1639532 Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Tested-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: correct function arguments to fix QNX compilationSourab Gupta2018-01-09
| | | | | | | | | | | | | | | The patch changes the function argument from 'int' to 'unsigned int' to fix the QNX compilation failures. Change-Id: Iaee7850d8310bea693996ac618b95252ca5d1b35 Signed-off-by: Sourab Gupta <sourabg@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1626397 Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> Reviewed-by: Deepak Nibade <dnibade@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Remove bare channel schedulingTerje Bergstrom2018-01-02
| | | | | | | | | | | | | | | | Remove scheduling IOCTL implementations for bare channels. Also removes code that constructs bare channels in runlist. Bug 1842197 Change-Id: I6e833b38e24a2f2c45c7993edf939d365eaf41f0 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1627326 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Remove support for channel eventsTerje Bergstrom2017-12-28
| | | | | | | | | | | | | | Remove support for events for bare channels. All users have already moved to TSGs and TSG events. Bug 1842197 Change-Id: Ib3ff68134ad9515ee761d0f0e19a3150a0b744ab Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1618906 Reviewed-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Remove support for bare channelsTerje Bergstrom2017-12-28
| | | | | | | | | | | | | | | Remove remaining support for bare channels. All users of bare channels have already moved to TSGs. Bug 1842197 Change-Id: I1ff12677253b160dac9bebe6925ad0839ea07cfc Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1618905 Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Disallow use of bare channelsTerje Bergstrom2017-12-12
| | | | | | | | | | | | | All channels need to now be wrapped in TSGs. Disallow use of bare channels by preventing creation of GPFIFO for them. Bug 1842197 Change-Id: Id0ebee4c590804b96c09f8951e35ba2680b596e7 Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1612697 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: cleanup uapi header includesDeepak Nibade2017-11-28
| | | | | | | | | | | | | | | | | | | | | With recent rework in nvgpu most of the <uapi/linux/nvgpu.h> includes are not needed so remove them Remove use of NVGPU_DBG_GPU_REG_OP_* in gk20a/gr_gk20a.c and use common definition instead Remove use of NVGPU_ALLOC_GPFIFO_FLAGS_REPLAYABLE_FAULTS_ENABLE in gp10b/fifo_gp10b.c by defining new common flag NVGPU_GPFIFO_FLAGS_REPLAYABLE_FAULTS_ENABLE and then parsing it in API nvgpu_gpfifo_user_flags_to_common_flags() Jira NVGPU-363 Change-Id: I8e653275ea3f443f24be7284d54f2115636aba3f Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1606108 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: move cycle state buffer handler to linuxDeepak Nibade2017-11-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | We use dma_buf pointer cyclestate_buffer_handler in common code But since this is linux specific, we need to move this out of common code and into linux specific code Move dma_buf pointer cyclestate_buffer_handler from common channel code to struct nvgpu_channel_linux Fix all pointer accesses to this handle Move gk20a_channel_free_cycle_stats_buffer() to ioctl_channel.c since it is mostly linux specific And since gk20a_channel_free_cycle_stats_buffer() needs to be called while closing the channel, call it from nvgpu_channel_close_linux() Jira NVGPU-397 Jira NVGPU-415 Change-Id: Ifb429e49b8f7a1c9e2bc757f3efdd50b28ceca1f Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1603909 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: move snapshot_client memory handling to linuxDeepak Nibade2017-11-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We right now store dmabuf fd and dma_buf pointer for gk20a_cs_snapshot_client But since dma_buf and all related APIs are linux specific, we need to remove them from common code and move them to linux specific code Add new linux specific structure gk20a_cs_snapshot_client_linux which includes struct gk20a_cs_snapshot_client and linux specific dma_buf pointer In gk20a_attach_cycle_stats_snapshot(), we first handle all dma_buf related operations and then call gr_gk20a_css_attach() Move gk20a_channel_free_cycle_stats_snapshot() to ioctl_channel.c In gk20a_channel_free_cycle_stats_snapshot(), we call gr_gk20a_css_detach() and then free up dma_buf in linux specific code We also need to call gk20a_channel_free_cycle_stats_snapshot() while closing the channel, so call it from linux specific nvgpu_channel_close_linux() Jira NVGPU-397 Jira NVGPU-415 Change-Id: Ida27240541f6adf31f28d7d7ee4f51651c6d3de2 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1603908 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: define error_notifiers in common codeDeepak Nibade2017-11-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | All the linux specific error_notifier codes are defined in linux specific header file <uapi/linux/nvgpu.h> and used in all the common driver But since they are defined in linux specific file, we need to move all the uses of those error_notifiers in linux specific code only Hence define new error_notifiers in include/nvgpu/error_notifier.h and use them in the common code Add new API nvgpu_error_notifier_to_channel_notifier() to convert common error_notifier of the form NVGPU_ERR_NOTIFIER_* to linux specific error notifier of the form NVGPU_CHANNEL_* Any future additions to error notifiers requires update to both the form of error notifiers Move all error notifier related metadata from channel_gk20a (common code) to linux specific structure nvgpu_channel_linux Update all accesses to this data from new structure instead of channel_gk20a Move and rename below APIs to linux specific file and declare them in error_notifier.h nvgpu_set_error_notifier_locked() nvgpu_set_error_notifier() nvgpu_is_error_notifier_set() Add below new API and use it in fifo_vgpu.c nvgpu_set_error_notifier_if_empty() Include <nvgpu/error_notifier.h> wherever new error_notifier codes are used NVGPU-426 Change-Id: Iaa5bfc150e6e9ec17d797d445c2d6407afe9f4bd Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1593361 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: use submit callback only in linux codeKonsta Holtta2017-11-22
| | | | | | | | | | | | | | | | | Move the implementation for channel job update callbacks that is based on Linux specific work_struct usage to Linux-specific code. This requires a bit of extra work for allocating OS-specific priv data for channels which is also done in this patch. The priv data will be used more when more OS-specific features are moved. Jira NVGPU-259 Change-Id: I24bc0148a827f375b56a1c96044685affc2d1e8c Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1589321 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: remove NVGPU_ALLOC_OBJ_FLAGS_* from common codeDeepak Nibade2017-11-10
| | | | | | | | | | | | | | | | | | | | In gr_gp10b_alloc_gr_ctx(), we use linux specific flags NVGPU_ALLOC_OBJ_FLAGS_* Since common code should be independent of linux specific code, define new flags NVGPU_OBJ_CTX_FLAGS_SUPPORT_* in common code and use them wherever needed Linux code will parse the user flags and send appropriate flags to g->ops.gr.alloc_obj_ctx() Also remove use of NVGPU_ALLOC_OBJ_FLAGS_LOCKBOOST_ZERO since this seems to be deadcode anyways Jira NVGPU-382 Change-Id: Id82efe0d46ddc3e2c063610025ea57f283bc3510 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1594452 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: remove NVGPU_ALLOC_GPFIFO_EX_FLAGS_* from common codeDeepak Nibade2017-11-10
| | | | | | | | | | | | | | | | | | | | | In gk20a_channel_alloc_gpfifo(), we use linux specific flags NVGPU_ALLOC_GPFIFO_EX_FLAGS_* Since common code should be independent of linux specific code, define new flags NVGPU_GPFIFO_FLAGS_SUPPORT_* in common code and use them in gk20a_channel_alloc_gpfifo() Linux code will parse the user flags and send appropriate flags to gk20a_channel_alloc_gpfifo() Jira NVGPU-381 Change-Id: Ibec51903b3407175fbba727208483b0dc36a5772 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1594422 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: support tuning per-ch deterministic optsKonsta Holtta2017-11-06
| | | | | | | | | | | | | | | | Add a new ioctl NVGPU_GPU_IOCTL_SET_DETERMINISTIC_OPTS to adjust deterministic options on a per-channel basis. Currently, the only supported option is to relax the no-railgating requirement on open deterministic channels. This also disallows submits on such channels, until the railgate option is reset. Bug 200327089 Change-Id: If4f0f51fd1d40ad7407d13638150d7402479aff0 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1554563 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: protect stacktrace.h include with configDeepak Nibade2017-11-06
| | | | | | | | | | | | | | | | struct stack_trace is protected with config GK20A_CHANNEL_REFCOUNT_TRACKING and hence protect linux/stacktrace.h header include in gk20a/channel_gk20a.h with same config Jira NVGPU-259 Change-Id: I365a4faa7eb071dd559e9b27fe03377dede7484d Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1592603 GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: move submit path to linuxDeepak Nibade2017-11-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Nvgpu submit path has a lot of dependency on Linux framework e.g. use of copy_from_user, use of structures defined in uapi/nvgpu headers, dma_buf_* calls for trace support etc Hence to keep common code independent of Linux code, move submit path to Linux directory Move below APIs to common/linux/channel.c trace_write_pushbuffer() trace_write_pushbuffer_range() gk20a_submit_prepare_syncs() gk20a_submit_append_priv_cmdbuf() gk20a_submit_append_gpfifo() gk20a_submit_channel_gpfifo() Move below APIs to common/linux/ce2.c gk20a_ce_execute_ops() Define gk20a_ce_execute_ops() in common/linux/ce2.c, and declare it in gk20a/ce2_gk20a.h since it is needed in common/mm code too Each OS needs to implement this API separately gk20a_channel_alloc_gpfifo() use sizeof(nvgpu_gpfifo) to get size of one gpfifo entry, but structure nvgpu_gpfifo is linux specific Define new nvgpu_get_gpfifo_entry_size() in linux specific code and use it in gk20a_channel_alloc_gpfifo() to get gpfifo entry size Each OS needs to implement this API separately Export some APIs from gk20a/ce2_gk20a.h and gk20a/channel_gk20a.h that are needed in linux code Jira NVGPU-259 Jira NVGPU-313 Change-Id: I360c6cb8ce4494b1e50c66af334a2a379f0d2dc4 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1586277 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: replace wait_queue_head_t with nvgpu_condDebarshi Dutta2017-10-16
| | | | | | | | | | | | | Replace existing usages of wait_queue_head_t with struct nvgpu_cond and using the corresponding APIs in order to reduce Linux dependencies in NVGPU. JIRA NVGPU-205 Change-Id: I85850369c3c47d3e1704e4171b1d172361842423 Signed-off-by: Debarshi Dutta <ddutta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1575778 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: protect stack_trace with configDeepak Nibade2017-10-13
| | | | | | | | | | | | | | | | We use struct stack_trace in struct channel_gk20a_ref_action But since channel_gk20a_ref_action is needed only if GK20A_CHANNEL_REFCOUNT_TRACKING is set, protect it with that config Jira NVGPU-259 Change-Id: I6b2d6f470bf924bb1ddfd31ba9968b56c63c2372 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1576929 Reviewed-by: Alex Waterman <alexw@nvidia.com> Reviewed-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> GVS: Gerrit_Virtual_Submit
* gpu: nvgpu: clean up channel open/release declaresDeepak Nibade2017-10-10
| | | | | | | | | | | | | | | | | | | Below APIs are already declared in ioctl_channel.h, and hence remove duplicate declaration from channel_gk20a.h gk20a_channel_open() gk20a_channel_ioctl() gk20a_channel_release() And move declaration of gk20a_channel_open_ioctl() from channel_gk20a.h to ioctl_channel.h Jira NVGPU-259 Change-Id: I46702ca481e41a19f92f4fe0169f95e31360abe0 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1573106 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: Change license for common files to MITTerje Bergstrom2017-09-26
| | | | | | | | | | | | Change license of OS independent source code files to MIT. JIRA NVGPU-218 Change-Id: I1474065f4b552112786974a16cdf076c5179540e Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1565880 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: fix channel unbind sequence from TSGDeepak Nibade2017-09-15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We right now remove a channel from TSG list and disable all the channels in TSG while removing a channel from TSG With this sequence if any one channel in TSG is closed, rest of the channels are set as timed out and cannot be used anymore We need to fix this sequence as below to allow removing a channel from active TSG so that rest of the channels can still be used - disable all channels of TSG - preempt TSG - check if CTX_RELOAD is set if support is available if CTX_RELOAD is set on channel, it should be moved to some other channel - check if FAULTED is set if support is available - if NEXT is set on channel then it means channel is still active print out an error in this case for the time being until properly handled - remove the channel from runlist - remove channel from TSG list - re-enable rest of the channels in TSG - clean up the channel (same as regular channels) Add below fifo operations to support checking channel status g->ops.fifo.tsg_verify_status_ctx_reload g->ops.fifo.tsg_verify_status_faulted Define ops.fifo.tsg_verify_status_ctx_reload operation for gm20b/gp10b/gp106 as gm20b_fifo_tsg_verify_status_ctx_reload() This API will check if channel to be released has CTX_RELOAD set, if yes CTX_RELOAD needs to be moved to some other channel in TSG Remove static from channel_gk20a_update_runlist() and export it Bug 200327095 Change-Id: I0dd4be7c7e0b9b759389ec12c5a148a4b919d3e2 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1560637 Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: Add wrapper over atomic_t and atomic64_tDebarshi Dutta2017-08-17
| | | | | | | | | | | | | | | | | - added wrapper structs nvgpu_atomic_t and nvgpu_atomic64_t over atomic_t and atomic64_t - added nvgpu_atomic_* and nvgpu_atomic64_* APIs to access the above wrappers. JIRA NVGPU-121 Change-Id: I61667bb0a84c2fc475365abb79bffb42b8b4786a Signed-off-by: Debarshi Dutta <ddutta@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1533044 Reviewed-by: svccoveritychecker <svccoveritychecker@nvidia.com> Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com> GVS: Gerrit_Virtual_Submit
* gpu: nvgpu: Fix gr ctx unmap logicAlex Waterman2017-07-27
| | | | | | | | | | | | | The GR context buffers were not being properly unmapped. The awkward VPR vs non-VPR context setup requires some extra checks when determining which nvgpu_mem is associated with what GPU VA (which are tracked separately in a different sized array). Change-Id: I4c7be1c5b7835aea4309a142df5b0bdfaae91e4c Signed-off-by: Alex Waterman <alexw@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1524689 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: add support for t19x tsg/channelseshendra Gadagottu2017-07-05
| | | | | | | | | | | | | | Required modifications to add t19x channel specific info and handle t19x tsg requests. Bug 1842197 Change-Id: I0f8bcce20edea8f2f9a01e5bf5a9e4181af54875 Signed-off-by: seshendra Gadagottu <sgadagottu@nvidia.com> Reviewed-on: https://git-master/r/1511144 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
* gpu: nvgpu: rename hw_chid to chidRichard Zhao2017-06-30
| | | | | | | | | | | | | hw_chid is a relative id for vgpu. For native it's same as hw id. Renaming it to chid to avoid confusing. Jira VFND-3796 Change-Id: I1c7924da1757330ace715a7c52ac61ec9dc7065c Signed-off-by: Richard Zhao <rizhao@nvidia.com> Reviewed-on: https://git-master/r/1509530 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: use time API in channel ref action debugKonsta Holtta2017-06-19
| | | | | | | | | | | | | Save the time using nvgpu_current_time_ms() instead of the Linux-specific jiffies counter. Jira NVGPU-83 Change-Id: I19b4296d8b64ddf52506144e77d151f668ff7838 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1503002 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
* gpu: nvgpu: hold power ref for deterministic channelsKonsta Holtta2017-06-14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | To support deterministic channels even with platforms where railgating is supported, have each deterministic-marked channel hold a power reference during their lifetime, and skip taking power refs for jobs in submit path for those. Previously, railgating blocked deterministic submits in general because of gk20a_busy()/gk20a_idle() calls in submit path possibly taking time and more significantly because the gpu may need turning on which takes a nondeterministic and long amount of time. As an exception, gk20a_do_idle() can still block deterministic submits until gk20a_do_unidle() is called. Add a rwsem to guard this. VPR resize needs do_idle, which conflicts with deterministic channels' requirement to keep the GPU on. This is documented in the ioctl header now. Make NVGPU_GPU_FLAGS_SUPPORT_DETERMINISTIC_SUBMIT_NO_JOBTRACKING always set in the gpu characteristics now that it's supported. The only thing left now blocking NVGPU_GPU_FLAGS_SUPPORT_DETERMINISTIC_SUBMIT_FULL is the sync framework. Make the channel debug dump show which channels are deterministic. Bug 200291300 Jira NVGPU-70 Change-Id: I47b6f3a8517cd6e4255f6ca2855e3dd912e4f5f3 Signed-off-by: Konsta Holtta <kholtta@nvidia.com> Reviewed-on: http://git-master/r/1483038 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
generated by cgit v1.2.2 (git 2.25.0) at 2024-07-04 22:24:09 -0400