nv-tegra-mirror/linux-nvgpu
1
0
Fork 0
mirror of git://nv-tegra.nvidia.com/linux-nvgpu.git synced 2025-12-24 02:22:34 +03:00
Code Issues Packages Projects Releases Wiki Activity

gpu: nvgpu: SWUD for nvhost-sync point unit

Browse Source 
This is adding Doxygen for nvhost-sync point unit.

Change-Id: I7a45f2a265beee479638bcd5c837c450ef8e986e
Signed-off-by: dinesh <dt@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2211714
Reviewed-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Ankur Kishore <ankkishore@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This commit is contained in:
dinesh
2019-10-04 19:42:14 +05:30
committed by Alex Waterman
parent da274f1fbc
commit 2240c543ca
1 changed files with 135 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

135
drivers/gpu/nvgpu/include/nvgpu/nvhost.h
View File

@@ -33,7 +33,39 @@ struct sync_pt;
struct sync_fence;
struct timespec;
/**
* @file Functions that initialize the sync points
* and describe other functionalities.
*/
/**
* @brief Initialzes the nvhost device for nvgpu.
*
* @param g [in] The GPU super structure.
*
* - Allocate memory for g.nvhost_dev.
* - Get the type of the hardware(t186/t194) by calling
* #NvTegraSysGetChipId().
* - Initialize the number of synpoints according to the
* associated hardware.
* - Allocate and initialize different fields associated with
* nvhost device.
*
* @return 0, if success.
* -ENOMEM, if it fails.
*/
int nvgpu_get_nvhost_dev(struct gk20a *g);
/**
* @brief Free the nvhost device.
*
* @param g [in] The GPU super structure.
*
* - Free the different fields of nvhost device initialized by
* #nvgpu_get_nvhost_dev().
*
* @return None.
*/
void nvgpu_free_nvhost_dev(struct gk20a *g);
#ifdef CONFIG_NVGPU_KERNEL_MODE_SUBMIT
@@ -69,18 +101,97 @@ const char *nvgpu_nvhost_syncpt_get_name(struct nvgpu_nvhost_dev *nvgpu_syncpt_d
#endif
/**
* @brief Increment the value of given sync point to the maximum value.
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param id [in] Sync point id.
*
* - Read the current value of the given sync point by calling
* #NvRmHost1xSyncpointRead().
* - Read the max value of the sync point set at allocation of the
* sync point.
* - If the max value is less than current, increment the syncpoint
* by the difference(max - current) using #NvRmHost1xSyncpointIncrement().
*
* @return None.
*/
void nvgpu_nvhost_syncpt_set_min_eq_max_ext(struct nvgpu_nvhost_dev *nvgpu_syncpt_dev,
u32 id);
/**
* @brief Read the maximum value of given sync point id.
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param id [in] Sync point id.
*
* - Read the max value of the given sync point by reading one of
* the max value array stored in nvhost sync point device structure.
*
* @return Maximum value of the sync point.
*/
u32 nvgpu_nvhost_syncpt_read_maxval(struct nvgpu_nvhost_dev *nvgpu_syncpt_dev,
u32 id);
/**
* @brief Set the value of given syncpoint to a value where all waiters of the
* sync point can be safely released.
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param id [in] Sync point id.
*
* - Read the current value of the sync point by #NvRmHost1xSyncpointRead().
* - Increment the value by 256. This is just to make the sync point safe
* where all waiters of the sync point can be safely released.
*
* @return None.
*/
void nvgpu_nvhost_syncpt_set_safe_state(
struct nvgpu_nvhost_dev *nvgpu_syncpt_dev, u32 id);
/**
* @brief Check the given sync point id is valid or not.
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param id [in] Sync point id.
*
* - Validate the sync point id.
*
* @return TRUE, If id is valid.
* FALSE, If id is not valid.
*/
bool nvgpu_nvhost_syncpt_is_valid_pt_ext(struct nvgpu_nvhost_dev *nvgpu_syncpt_dev,
u32 id);
/**
* @brief Free the sync point created by #nvgpu_nvhost_get_syncpt_client_managed().
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param id [in] Sync point id.
*
* - Check the validity of the given sync point.
* - Free the fields allocated by #nvgpu_nvhost_get_syncpt_client_managed().
*
* @return None.
*/
void nvgpu_nvhost_syncpt_put_ref_ext(struct nvgpu_nvhost_dev *nvgpu_syncpt_dev,
u32 id);
/**
* @brief Allocate a sync point managed by a client.
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param name [in] Name of the sync point.
*
* - Call #nvgpu_nvhost_allocate_syncpoint() to allocate sync point.
* -- nvgpu_nvhost_allocate_syncpoint() will do the following
* - Call #NvRmHost1xSyncpointAllocate() to allocate sync point.
* - Call #NvRmHost1xSyncpointGetId() to get the ID.
* - Call #NvRmHost1xWaiterAllocate() to get waiter handle if needed.
* - Store the above datas in to the nvgpu nvhost device.
*
* @return Sync point id allocated.
*/
u32 nvgpu_nvhost_get_syncpt_client_managed(struct nvgpu_nvhost_dev *nvgpu_syncpt_dev,
const char *syncpt_name);
@@ -96,9 +207,33 @@ struct sync_fence *nvgpu_nvhost_sync_create_fence(
#endif /* CONFIG_SYNC */
#ifdef CONFIG_TEGRA_T19X_GRHOST
/**
* @brief Initializes the address and size of memory mapped
* sync point unit region(MSS).
*
* @param nvgpu_syncpt_dev [in] Sync point device.
* @param base [out] Base address where it mapped.
* @param size [out] Size of the mapping.
*
* - Retrieve the value of base and size from the given
* sync point device.
*
* @return Zero.
*/
int nvgpu_nvhost_syncpt_unit_interface_get_aperture(
struct nvgpu_nvhost_dev *nvgpu_syncpt_dev,
u64 *base, size_t *size);
/**
* @brief Get offset of the sync point from MSS aperture base.
*
* @param syncpt_id [in] Sync point id.
*
* - Multiply the id with 0x1000.
*
* @return Sync point offset.
*/
u32 nvgpu_nvhost_syncpt_unit_interface_get_byte_offset(u32 syncpt_id);
int nvgpu_nvhost_syncpt_init(struct gk20a *g);
#else