| Commit message (Collapse) | Author | Age |
... | |
|
|
|
|
|
|
|
|
| |
Add gp10b PROD value for FE_GO_IDLE_TIMEOUT. Use the PROD value
written in gk20a_init_gr_setup_hw() instead of hard coding here.
Change-Id: If3bd981c1c0d9cc8ad19c21c220b7de81fdb529e
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/813959
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We used to allocate 1.5x buffer size. This leads to memory waste, as
we do not set the CB size via SW methods anymore.
Bug 1686189
Change-Id: I45cbdeadc154f59b65138f99f50a72d97511cb78
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/801865
(cherry picked from commit 791f2fe03d16521206649ab90498443e91e284e2)
Reviewed-on: http://git-master/r/815683
|
|
|
|
|
|
|
|
|
|
|
|
| |
Handle beta CB sizing differences for GfxP versus WFI channels.
Bug 1686189
Change-Id: Icc421eeb8305f7e4156a74c957662f19504ddad7
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/801533
(cherry picked from commit 95b9ae4e5f3c29fdb97567d846b9d2139f1a8ec4)
Reviewed-on: http://git-master/r/815682
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Spill buffer size is in chunks of 256B. Multiply the size by
granularity to get the size in bytes.
Bug 1686189
Change-Id: I0462293668322645bd1eab190c12faaeb6c316c1
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/801344
(cherry picked from commit 4bf6de7d9c9014a9eaeff56b19437d1841d7cfb0)
Reviewed-on: http://git-master/r/815680
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If pagepool size equals max we should use zero. Add the comparison
to do that.
Bug 1686189
Change-Id: I15bd43663550b1089a726c0256b89f849c193e21
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/801526
(cherry picked from commit 9d89ea5ba345b19d2cff86130ba9d3c4c5f07e6e)
Reviewed-on: http://git-master/r/815681
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug 200137963
Change-Id: Ibd09b206620e6d6826586bb40e1125fc178dd8e4
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: http://git-master/r/797151
(cherry picked from commit 343c4704564f4b4f22a943a94e66d2c83f63a28f)
Reviewed-on: http://git-master/r/808241
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Implement method for setting the coalesce buffer size at runtime.
Bug 1681992
Change-Id: Ice6c00a27f642c2d68d6cd0e30c12df2e48f5374
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/802366
(cherry picked from commit bd763bc8a16b80ccc8f79b2229eccf2fe2417611)
Reviewed-on: http://git-master/r/808239
GVS: Gerrit_Virtual_Submit
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Enabling cyclestats and cyclestats snapshot support for gp10b (t186) devices.
Bug 1674079
Change-Id: I2e14801de3c61d180630bb9dcd2c607749814893
Signed-off-by: Leonid Moiseichuk <lmoiseichuk@nvidia.com>
Reviewed-on: http://git-master/r/792953
Reviewed-on: http://git-master/r/806190
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We program the default steady state beta CB size. The default is
for deep binning, but we've disabled deep binning. As result steady
state CB size was left too high.
Bug 1683535
Change-Id: I17029078d9c83e55eec6faacfc83c6d812f8c3c0
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/795306
Reviewed-on: http://git-master/r/806189
|
|
|
|
|
|
|
|
|
|
| |
Program correct CB size for GfxP channels. We were accidentally
using the context image size.
Change-Id: I273215256e41e89b7d76f3294a73641804beeb79
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/792713
Reviewed-on: http://git-master/r/806188
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add debugfs switch to force cilp and gfx preemption
Add debugfs switch to dump context switch stats on channel
destruction.
bug 1525327
bug 1581799
Signed-off-by: Kirill Artamonov <kartamonov@nvidia.com>
Change-Id: I7d0558cc325ce655411388ea66ad982101f2fe66
Reviewed-on: http://git-master/r/794976
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/677231
|
|
|
|
|
|
|
|
|
| |
Disable deep binning by default.
Change-Id: I75da95984ac314015c6927e099a3eaa37fcc26fc
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/790403
Reviewed-on: http://git-master/r/806186
|
|
|
|
|
|
|
|
|
| |
Bug 1678603
Change-Id: Ib8fb09dace864567b1ce574c216a584831723684
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/790402
Reviewed-on: http://git-master/r/806185
|
|
|
|
|
|
|
|
|
| |
Bug 1642669
Change-Id: I683338256b7f2a165a7933aa59de510eb109ea6f
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/755150
Reviewed-by: Automatic_Commit_Validation_User
|
|
|
|
|
|
|
|
|
|
| |
Set address check mode for SM.
Bug 1625763
Change-Id: I5ddf8334673b414956e57c55aaa5be1a9f9aeaf1
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/752139
|
|
|
|
|
|
|
|
|
|
|
| |
Alpha and beta sizes need to be clipped to a maximum value. For
alpha CB we were using beta size in clipping, and for both we were
not using number of TPCs to determine the max value.
Change-Id: I0c925464ba4c9f575e6e59dd5ba7759aa1cb6381
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/752667
Reviewed-by: Automatic_Commit_Validation_User
|
|
|
|
|
|
|
|
|
| |
Some fields have different widths, so duplicate the code to program
global bundle CB.
Change-Id: Ib6af5abf3e90dfa1bcda2fbc6b97ad1031e6ab16
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/752635
|
|
|
|
|
|
|
|
| |
Program CB base to new gp10b registers.
Change-Id: I1ab39a487dade58d3a024fb1aba1af5c878f31bb
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/752634
|
|
|
|
|
|
|
|
|
|
| |
When allocating betacb for a GfxP channel, add both alpha and beta
cb sizes together.
Change-Id: I8cef62f6272bfb3b5e9a3835a51590e5eb91dc92
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/752633
Reviewed-by: Automatic_Commit_Validation_User
|
|
|
|
|
|
|
|
|
|
|
|
| |
Expose CILP and GFXP flags to user space ioctl
NVGPU_IOCTL_CHANNEL_ALLOC_OBJ_CTX.
Bug 200111328
Change-Id: I10931db2babd3222e308fd491824d95204355ff3
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/748932
GVS: Gerrit_Virtual_Submit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ZBC is safe to update and GPU is safe to rail gate when units are
in preempted or empty state. Idle may never be reached in case of
graphics preemption, so relax the ZBC update wait condition.
Bug 1640378
Change-Id: I40c59e9af22a7a30b777c6b9f87e69d130042e44
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/745655
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
|
|
|
|
|
|
|
|
|
|
| |
Calculate GFXP attrib cb buffer size from the global buffer size.
Bug 1628352
Change-Id: If4edfbf5700334b791dbf8e5cf38fd0208ee7fa1
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/735717
|
|
|
|
|
|
|
|
|
|
|
| |
If betacb size has been given via debugfs, use that instead of the
calculated number.
Bug 1628352
Change-Id: I8c68c27a2bfdd7f013776734ef846377a89b0033
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/733332
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add a missing bitmask for clearing existing bits before setting a new
value, and shift the value the correct amount. Also format register
needs to be rounded down.
Bug 200087330
Change-Id: I39051be7eb68327fc010495f0c16c879447c8e4c
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/726265
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fix sparse warnings of below type by making necessary
symbols static:
warning: symbol '<symbol>' was not declared. Should it be static?
Bug 200088648
Change-Id: Ic20ef3eb73dcbfe5f13506b5afa629c3e1db59d0
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: http://git-master/r/728012
GVS: Gerrit_Virtual_Submit
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
|
|
|
|
|
|
|
|
|
|
| |
Program steady state CB size to be the HW default.
Bug 1626065
Change-Id: If0bdc5a649f307b6adab4e914a6201222b8453f8
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/725106
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Move the fifo engine activity disabling and wait-for-idle from the
lowest-level functions higher, into the ioctl path of zbc operations, so
that the sw initialization path wouldn't call them. During the init
path, the disable isn't necessary, and the code path could result in a
deadlock in the fifo runlist mutex.
Change-Id: I56e73204e288331165358fc9856390f1eb724488
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: http://git-master/r/715196
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
|
|
|
|
|
|
|
|
|
|
| |
Offset for preemption pointer was calculated incorrectly.
Bug 1617214
Change-Id: I9c1a9ae24dcd523f4ae17eae0a5b07831839fadb
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/716528
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add below exception registers to GR dump :
NV_PGRAPH_PRI_BE0_BECS_BE_EXCEPTION
NV_PGRAPH_PRI_BE0_BECS_BE_EXCEPTION_EN
NV_PGRAPH_PRI_GPC0_GPCCS_GPC_EXCEPTION
NV_PGRAPH_PRI_GPC0_GPCCS_GPC_EXCEPTION_EN
NV_PGRAPH_PRI_GPC0_TPC0_TPCCS_TPC_EXCEPTION
NV_PGRAPH_PRI_GPC0_TPC0_TPCCS_TPC_EXCEPTION_EN
Bug 200078514
Change-Id: I2400e360fea0b3bdcdf5f3dd6ef250867fb191e6
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: http://git-master/r/712481
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixed incorrectly encoded pointer and size.
bug 1525327
bug 1581799
Change-Id: Ie6e94e47c3b11e9d9aa63a70b61e6e89f69e971b
Signed-off-by: Kirill Artamonov <kartamonov@nvidia.com>
Reviewed-on: http://git-master/r/713209
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
|
|
|
|
|
|
| |
Change-Id: I86de27309ebecd038a7b32c6f86d87ce0156eb14
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/709867
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add function pointer gr_gp10b_dump_gr_status_regs()
which will enable dumping GR status registers for gp10b
Bug 200062436
Change-Id: Iaecc2f9c9364232079bb03e114f68550bd035372
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: http://git-master/r/678832
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
bug 200078367
Change-Id: I9a68e988fa7921276e334c75afa5ee4b15aab464
Signed-off-by: Vijayakumar <vsubbu@nvidia.com>
Reviewed-on: http://git-master/r/707313
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
|
|
|
|
|
|
|
|
|
|
|
| |
Allow enabling CILP for compute. Set CTA by default.
Bug 1517461
Change-Id: I85cc931b810afb3ee6116de1200d01b52e1bc29e
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/661298
GVS: Gerrit_Virtual_Submit
|
|
|
|
|
|
|
|
|
|
|
| |
Implement context creation for gp10b. GfxP contexts need per channel
buffers.
Bug 1517461
Change-Id: Ifecb59002f89f0407457730a35bfb3fe988b907a
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/660236
|
|
|
|
|
|
|
|
|
| |
Change order of alpha & attribute buffers in CB. The new order
follows RM.
Change-Id: I2b24daa46055b3bd667a1026c282f74d56882623
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/657907
|
|
|
|
|
|
|
|
|
|
|
| |
Program CB sizes.
Bug 1567274
Change-Id: Idc88f69b70e85bf950af852a9ca80a328d95883f
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/654097
GVS: Gerrit_Virtual_Submit
|
|
|
|
|
|
|
|
| |
Bug 1567274
Change-Id: I6d760eca7ac0931847f9a04a9d4a408519ade511
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/654098
|
|
|
|
|
|
|
|
|
|
| |
In gp10b we need to limit global context buffer size, and it needs
to be 128b aligned.
Change-Id: I51570e2457a374c09be4d611e683ae30917f9fc0
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/657911
GVS: Gerrit_Virtual_Submit
|
|
|
|
|
|
|
|
|
| |
Bug 1567274
Change-Id: I4369458d3af0c4da32af8a5881c8fe60b11f7632
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/606932
Reviewed-by: Automatic_Commit_Validation_User
|
|
|
|
|
|
|
|
| |
Bug 1567274
Change-Id: Ife98ae512c62bd26450e59338719c7a10635b5dd
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/601108
|
|
|
|
|
|
|
|
|
| |
Bug 1570662
Change-Id: Icb7e90b1216acfd19bb3027dc9e9844eb08c99d9
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/592101
GVS: Gerrit_Virtual_Submit
|
|
This change adds gp10b to the nvgpu build as
well as enabling CMA for buffer allocation.
Change-Id: Id3d45ad6ffdab14120395952e68b285dd7364c76
Signed-off-by: Ken Adams <kadams@nvidia.com>
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/553324
GVS: Gerrit_Virtual_Submit
|