nvgpu.git/include, branch gpu-paging

gpu-paging: Allow for more than one buffer to be swapped at a time

2022-06-03T19:41:42+00:00

This uses a very primitive linear disk sector allocation scheme.
Sectors are only reused when userspace resets assignment to 0 with
an NVGPU_AS_IOCTL_SWAP_RESET ioctl (which invalidates all current
swap buffers).

This sector assignment scheme is sufficient for use in a TimeWall-
like system, where all allocations are assumed to be static after
after task system release. This is not suitable for a system with
dynamic allocations, unless userspace manually resets swap state
regularly (benchmarks run a reset at start for example).

Support for dynamic allocations is on the backlog.

No significant speed impact.

Benchmarks, 100 iters, after:
gpu_paging_speed, write: 186.0ms +/- 3.51
gpu_paging_speed, read: 162.7ms +/- 2.58
gpu_paging_overhead_speed, write start: 35.4ms +/- 4.47
gpu_paging_overhead_speed, write finish: 3.3ms +/- 0.18
gpu_paging_overhead_speed, read start: 69.8ms +/- 6.42
gpu_paging_overhead_speed, read finish: 43.2ms +/- 0.91

gpu-paging: Split swap in/out to prepare for async support.

2022-05-30T16:19:42+00:00

gpu-paging: Initial working implementation

2022-05-25T01:11:59+00:00

Supports synchronous page out or in of a specific buffer.

Includes fast reverse struct mapped_buf lookup.

Requires initial set of changes to nvmap as well.

gpu: nvgpu: remove channel cycle stats ioctls

2020-07-27T21:55:13+00:00

Cycle stats and cycle stats snapshot ioctls have been moved to
debug node. Removing channel ioctls.

Bug 2660206
Bug 220464613

Change-Id: I3aecdf4a8310eeb38de2de5ac076048891afe436
Signed-off-by: Thomas Fleury 
Reviewed-on: https://git-master.nvidia.com/r/2030992
(cherry picked from commit f20424ea6a7c6fcf977630e3e95d9e78418f13b8)
Signed-off-by: Gagan Grover 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2092020
Reviewed-by: automaticguardword 
Reviewed-by: Phoenix Jung 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit
Tested-by: Peter Daifuku 
Tested-by: mobile promotions

gpu: nvgpu: add cycle stats to debugger node

2020-07-27T21:55:01+00:00

Add NVGPU_DBG_GPU_IOCTL_CYCLE_STATS to debugger node, to
install/uninstall a buffer for cycle stats.

Add NVGPU_DBG_GPU_IOCTL_CYCLE_STATS_SNAPSHOT to debugger
node, to attach/flush/detach a buffer for Mode-E streamout.

Those ioctls will apply to the first channel in the debug session.

Bug 2660206
Bug 200464613

Change-Id: I0b96d9a07c016690140292fa5886fda545697ee6
Signed-off-by: Thomas Fleury 
Reviewed-on: https://git-master.nvidia.com/r/2002060
(cherry picked from commit 90b0bf98ac01d7fa24c40f6a1f20bfe5fa481d36)
Signed-off-by: Gagan Grover 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2092008
Reviewed-by: automaticguardword 
Reviewed-by: Phoenix Jung 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit
Tested-by: Peter Daifuku 
Tested-by: mobile promotions

nvgpu: include: Fix NVGPU_TSG_IOCTL_READ_SINGLE_SM_ERROR_STATE definition

2020-02-17T22:39:10+00:00

IOCTL definition specifies _IOR whereas _IOWR is required

Bug 200412642

Change-Id: I1093362ea621ee507d19236b859b7defb6dfa090
Signed-off-by: Anup Mahindre 
Reviewed-on: https://git-master.nvidia.com/r/1920071
(cherry picked from commit 79af49f86b90a80e20509a3b008440e9b9680cf9)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2287838
Reviewed-by: Deepak Nibade 
Reviewed-by: Sami Kiminki 
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit
Tested-by: Thomas Fleury 
Tested-by: mobile promotions

gpu: nvgpu: add SET_CTX_MMU_DEBUG_MODE ioctl

2020-01-30T07:41:45+00:00

Added NVGPU_DBG_GPU_IOCTL_SET_CTX_MMU_DEBUG_MODE ioctl to set MMU
debug mode for a given context.

Added gr.set_mmu_debug_mode HAL to change NV_PGPC_PRI_MMU_DEBUG_CTRL
for a given channel. HAL implementation for native case is
gm20b_gr_set_mmu_debug_mode. It internally uses regops, which directly
writes to the register if the context is resident, or writes to
gr context otherwise.

Added NVGPU_SUPPORT_SET_CTX_MMU_DEBUG_MODE to enable the feature.
NV_PGPC_PRI_MMU_DEBUG_CTRL has to be context switched in FECS ucode,
so the feature is only enabled on TU104 for now.

Bug 2515097
But 2713590

Change-Id: Ib4efaf06fc47a8539b4474f94c68c20ce225263f
Signed-off-by: Thomas Fleury 
Reviewed-on: https://git-master.nvidia.com/r/2110720
(cherry-picked from commit af2ccb811d3de06f052b1dee39bd9ffa863ac8ce)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2208767
Reviewed-by: Kajetan Dutka 
Reviewed-by: Alex Waterman 
Reviewed-by: Winnie Hsu 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit
Tested-by: Kajetan Dutka 
Tested-by: mobile promotions

gpu: nvgpu: add platform atomic support

2020-01-08T16:35:30+00:00

Add new variable in nvgpu_as_map_buffer_ex_args for app
to specify the platform atomic support for the page.
When platform atomic attribute flag is set, pte memory
aperture is set to be coherent type.

renamed nvgpu_aperture_mask_coh -> nvgpu_aperture_mask_raw
function.

bug 200580236

Change-Id: I18266724dafdc8dfd96a0711f23cf08e23682afc
Signed-off-by: Vinod G 
Reviewed-on: https://git-master.nvidia.com/r/2012679
(cherry picked from commit 9e0a9004b71f92b7713fd3b30141b0d9d4cfa2c6)
Signed-off-by: Lakshmanan M 
Reviewed-on: https://git-master.nvidia.com/r/2274914
Reviewed-by: Deepak Nibade 
GVS: Gerrit_Virtual_Submit
Tested-by: Sreeniketh H 
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
Tested-by: mobile promotions

gpu: nvgpu: mark usermode submit supported for gv11b

2019-08-15T07:59:04+00:00

Mark usermode submit supported in gv11b and add the characteristics flag
to expose the capability to userspace.

Bug 200145225
Bug 200541476

Change-Id: Id9dcb0c71c020bd509fbdbffb94a756c69377f20
Signed-off-by: Konsta Holtta 
Reviewed-on: https://git-master.nvidia.com/r/1795822
Reviewed-by: Alex Waterman 
Reviewed-by: Terje Bergstrom 
Reviewed-by: Vijayakumar Subbu 
Signed-off-by: Debarshi Dutta 
(cherry picked from commit 37659f5c8e0571655178c50a6296b68e3ebdc4cb
in rel-32)
Reviewed-on: https://git-master.nvidia.com/r/2170604
GVS: Gerrit_Virtual_Submit
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
Tested-by: mobile promotions

gpu: nvgpu: support usermode submit buffers

2019-08-15T07:58:54+00:00

Import userd and gpfifo buffers from userspace if provided via
NVGPU_IOCTL_CHANNEL_ALLOC_GPFIFO_EX. Also supply the work submit token
(i.e., the hw channel id) to userspace.

To keep the buffers alive, store their dmabuf and attachment/sgt handles
in nvgpu_channel_linux. Our nvgpu_mem doesn't provide such data for
buffers that are mainly in kernel use. The buffers are freed via a new
API in the os_channel interface.

Fix a bug in gk20a_channel_free_usermode_buffers: also unmap the
usermode gpfifo buffer.

Bug 200145225
Bug 200541476

Change-Id: I8416af7085c91b044ac8ccd9faa38e2a6d0c3946
Signed-off-by: Konsta Holtta 
Reviewed-on: https://git-master.nvidia.com/r/1795821
Signed-off-by: Debarshi Dutta 
(cherry picked from commit 99b1c6dcdf328efcfe47338ad1b71a114ab7f272
in dev-main)
Reviewed-on: https://git-master.nvidia.com/r/2170603
GVS: Gerrit_Virtual_Submit
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
Tested-by: mobile promotions