mirror of
git://nv-tegra.nvidia.com/linux-nvgpu.git
synced 2025-12-22 09:12:24 +03:00
Most of the Orin chip specific code is compiled out of safety build with CONFIG_NVGPU_NON_FUSA and CONFIG_NVGPU_HAL_NON_FUSA. Remove the config protection from Orin/GA10B specific code. Currently all code is enabled. Code not required in safety will be compiled out later in separate activity. Other noteworthy changes in this patch related to safety build: - In ga10b_ce_request_idle(), add a log print to dump num_pce so that compiler does not complain about unused variable num_pce. - In ga10b_fifo_ctxsw_timeout_isr(), protect variables active_eng_id and recover under CONFIG_NVGPU_KERNEL_MODE_SUBMIT to fix compilation errors of unused variables. - Compile out HAL gops.pbdma.force_ce_split() from safety since this HAL is GA100 specific and not required for GA10B. - Compile out gr_ga100_process_context_buffer_priv_segment() with CONFIG_NVGPU_DEBUGGER. - Compile out VAB support with CONFIG_NVGPU_HAL_NON_FUSA. - In ga10b_gr_intr_handle_sw_method(), protect left_shift_by_2 variable with appropriate configs to fix unused variable compilation error. - In ga10b_intr_isr_stall_host2soc_3(), compile ELPG function calls with CONFIG_NVGPU_POWER_PG. - In ga10b_pmu_handle_swgen1_irq(), move whole function body under CONFIG_NVGPU_FALCON_DEBUG to fix unused variable compilation errors. - Add below TU104 specific files in safety build since some of the code in those files is required for GA10B. Unnecessary code will be compiled out later on. hal/gr/init/gr_init_tu104.c hal/class/class_tu104.c hal/mc/mc_tu104.c hal/fifo/usermode_tu104.c hal/gr/falcon/gr_falcon_tu104.c - Compile out GA10B specific debugger/profiler related files from safety build. - Disable CONFIG_NVGPU_FALCON_DEBUG from safety debug build temporarily to work around compilation errors seen with keeping this config enabled. Config will be re-enabled in safety debug build later. Jira NVGPU-7276 Change-Id: I35f2489830ac083d52504ca411c3f1d96e72fc48 Signed-off-by: Deepak Nibade <dnibade@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2627048 Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
295 lines
9.8 KiB
C
295 lines
9.8 KiB
C
/*
|
|
* Copyright (c) 2020-2021, NVIDIA CORPORATION. All rights reserved.
|
|
*
|
|
* Permission is hereby granted, free of charge, to any person obtaining a
|
|
* copy of this software and associated documentation files (the "Software"),
|
|
* to deal in the Software without restriction, including without limitation
|
|
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
|
|
* and/or sell copies of the Software, and to permit persons to whom the
|
|
* Software is furnished to do so, subject to the following conditions:
|
|
*
|
|
* The above copyright notice and this permission notice shall be included in
|
|
* all copies or substantial portions of the Software.
|
|
*
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
|
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
|
|
* DEALINGS IN THE SOFTWARE.
|
|
*/
|
|
#ifndef NVGPU_GOPS_MC_H
|
|
#define NVGPU_GOPS_MC_H
|
|
|
|
#include <nvgpu/types.h>
|
|
#include <nvgpu/mc.h>
|
|
#include <nvgpu/engines.h>
|
|
|
|
/**
|
|
* @file
|
|
*
|
|
* MC HAL interface.
|
|
*/
|
|
struct gk20a;
|
|
struct nvgpu_device;
|
|
|
|
/**
|
|
* MC HAL operations.
|
|
*
|
|
* @see gpu_ops.
|
|
*/
|
|
struct gops_mc {
|
|
/**
|
|
* @brief Get the GPU architecture, implementation and revision.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
* @param arch [out] GPU architecture level pointer. Can be
|
|
* NULL if not needed by caller.
|
|
* @param impl [out] GPU architecture pointer. Can be
|
|
* NULL if not needed by caller.
|
|
* @param rev [out] Chip revision pointer. Can be
|
|
* NULL if not needed by the caller.
|
|
*
|
|
* This function is invoked to get the GPU architecture, implementation
|
|
* and revision level of the GPU chip before #nvgpu_finalize_poweron.
|
|
* These values are used for chip specific SW/HW handling in the
|
|
* driver.
|
|
*
|
|
* Steps:
|
|
* - Read the register mc_boot_0_r().
|
|
* - Architecture ID is placed in \arch
|
|
* - GPU implementation ID is placed in \a impl
|
|
* - Chip revision is placed in \a rev
|
|
*
|
|
* @return value of mc_boot_0_r().
|
|
*/
|
|
u32 (*get_chip_details)(struct gk20a *g,
|
|
u32 *arch, u32 *impl, u32 *rev);
|
|
|
|
/**
|
|
* @brief Read the stalling interrupts status register.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
*
|
|
* This function is invoked to get stalling interrupts reported
|
|
* by the GPU before invoking the ISR.
|
|
*
|
|
* Steps:
|
|
* - Read and return the value of register
|
|
* mc_intr_r(#NVGPU_CIC_INTR_STALLING).
|
|
*
|
|
* @return value read from mc_intr_r(#NVGPU_CIC_INTR_STALLING).
|
|
*/
|
|
u32 (*intr_stall)(struct gk20a *g);
|
|
|
|
/**
|
|
* @brief Interrupt Service Routine (ISR) for handling the stalling
|
|
* interrupts.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
*
|
|
* This function is called by OS interrupt unit on receiving
|
|
* stalling interrupt for servicing it.
|
|
*
|
|
* Steps:
|
|
* - Read mc_intr_r(#NVGPU_CIC_INTR_STALLING) register to get the
|
|
* stalling interrupts reported.
|
|
* - For the FIFO engines with pending interrupt invoke corresponding
|
|
* handlers.
|
|
* - Invoke g->ops.gr.intr.stall_isr if GR interrupt is pending.
|
|
* - Invoke g->ops.ce.isr_stall if CE interrupt is pending.
|
|
* - For other units with pending interrupt invoke corresponding
|
|
* handlers.
|
|
* - Invoke g->ops.fb.intr.isr if HUB interrupt is pending, determined
|
|
* by calling g->ops.mc.is_intr_hub_pending.
|
|
* - Invoke g->ops.fifo.intr_0_isr if FIFO interrupt is pending. The
|
|
* FIFO interrupt bit in mc_intr_r(#NVGPU_CIC_INTR_STALLING) is
|
|
* mc_intr_pfifo_pending_f.
|
|
* - Invoke g->ops.pmu.pmu_isr if PMU interrupt is pending.
|
|
* The PMU interrupt bit in mc_intr_r(#NVGPU_CIC_INTR_STALLING)
|
|
* is mc_intr_pmu_pending_f.
|
|
* - Invoke g->ops.priv_ring.isr if PRIV_RING interrupt is pending.
|
|
* The PRIV_RING interrupt bit in mc_intr_r(#NVGPU_CIC_INTR_STALLING)
|
|
* is mc_intr_priv_ring_pending_f.
|
|
* - Invoke g->ops.mc.ltc_isr if LTC interrupt is pending. The
|
|
* LTC interrupt bit in mc_intr_r(#NVGPU_CIC_INTR_STALLING) is
|
|
* mc_intr_ltc_pending_f.
|
|
* - Invoke g->ops.bus.isr if BUS interrupt is pending. The
|
|
* BUS interrupt bit in mc_intr_r(#NVGPU_CIC_INTR_STALLING) is
|
|
* mc_intr_pbus_pending_f.
|
|
*/
|
|
void (*isr_stall)(struct gk20a *g);
|
|
|
|
/**
|
|
* @brief Read the non-stalling interrupts status register.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
*
|
|
* This function is invoked to get the non-stalling interrupts reported
|
|
* by the GPU before invoking the ISR.
|
|
*
|
|
* Steps:
|
|
* - Read and return the value of the register
|
|
* mc_intr_r(#NVGPU_CIC_INTR_NONSTALLING).
|
|
*
|
|
* @return value read from mc_intr_r(#NVGPU_CIC_INTR_NONSTALLING).
|
|
*/
|
|
u32 (*intr_nonstall)(struct gk20a *g);
|
|
|
|
/**
|
|
* @brief Interrupt Service Routine (ISR) for handling the non-stalling
|
|
* interrupts.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
*
|
|
* This function is called by OS interrupt unit on receiving
|
|
* non-stalling interrupt for servicing it.
|
|
*
|
|
* Steps:
|
|
* - Read mc_intr_r(#NVGPU_CIC_INTR_NONSTALLING) register to get the
|
|
* non-stalling interrupts reported.
|
|
* - Invoke g->ops.fifo.intr_1_isr if FIFO non-stalling interrupt
|
|
* is pending, determined by calling mc_intr_pfifo_pending_f.
|
|
* - For the FIFO engines with pending interrupt invoke corresponding
|
|
* handlers.
|
|
* - Invoke g->ops.gr.intr.nonstall_isr if GR interrupt is pending.
|
|
* - Invoke g->ops.ce.isr_nonstall if CE interrupt is pending.
|
|
*
|
|
* @return bitmask of operations that are executed on non-stall
|
|
* workqueue.
|
|
*/
|
|
u32 (*isr_nonstall)(struct gk20a *g);
|
|
|
|
/**
|
|
* @brief Check if stalling or engine interrupts are pending.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
* @param engine_id [in] Active engine id.
|
|
* - Min: 0
|
|
* - Max: NV_HOST_NUM_ENGINES
|
|
* @param eng_intr_pending [out] Pointer to get pending
|
|
* interrupt.
|
|
*
|
|
* This function is invoked while polling for preempt completion.
|
|
*
|
|
* Steps:
|
|
* - Read mc_intr_r(#NVGPU_CIC_INTR_STALLING) register to get
|
|
* the interrupts reported.
|
|
* - Get the engine interrupt mask corresponding to \a engine_id.
|
|
* - Check if the bits for engine interrupt mask are set in the
|
|
* mc_intr_r(#NVGPU_CIC_INTR_STALLING) register by AND'ing values
|
|
* read in above two steps. Store the result in \a eng_intr_pending.
|
|
* - Initialize the stalling interrupt mask with bitmask for FIFO, HUB,
|
|
* PRIV_RING, PBUS, LTC unit interrupts.
|
|
* - Return true if bits from above stalling interrupt mask or the
|
|
* engine interrupt mask are set in the
|
|
* mc_intr_r(#NVGPU_CIC_INTR_STALLING) register. Else, return false.
|
|
*
|
|
* @return true if stalling or engine interrupt is pending, else false.
|
|
*/
|
|
bool (*is_stall_and_eng_intr_pending)(struct gk20a *g,
|
|
u32 engine_id, u32 *eng_intr_pending);
|
|
|
|
/**
|
|
* @brief Interrupt Service Routine (ISR) for handling the Level Two
|
|
* Cache (LTC) interrupts.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
*
|
|
* This function is invoked to handle the LTC interrupts from
|
|
* #isr_stall.
|
|
*
|
|
* Steps:
|
|
* - Read mc_intr_ltc_r register to get the interrupts status for LTCs.
|
|
* - For each ltc from index 0 to nvgpu_ltc_get_ltc_count(\a g)
|
|
* - If interrupt bitmask is set in the interrupts status register
|
|
* - Invoke g->ops.ltc.intr.isr.
|
|
*/
|
|
void (*ltc_isr)(struct gk20a *g);
|
|
|
|
/** @cond DOXYGEN_SHOULD_SKIP_THIS */
|
|
|
|
void (*intr_mask)(struct gk20a *g);
|
|
|
|
void (*intr_enable)(struct gk20a *g);
|
|
|
|
void (*intr_stall_unit_config)(struct gk20a *g, u32 unit,
|
|
bool enable);
|
|
|
|
void (*intr_nonstall_unit_config)(struct gk20a *g, u32 unit,
|
|
bool enable);
|
|
|
|
void (*intr_stall_pause)(struct gk20a *g);
|
|
|
|
void (*intr_stall_resume)(struct gk20a *g);
|
|
|
|
void (*intr_nonstall_pause)(struct gk20a *g);
|
|
|
|
void (*intr_nonstall_resume)(struct gk20a *g);
|
|
|
|
int (*enable_units)(struct gk20a *g, u32 units, bool enable);
|
|
|
|
int (*enable_dev)(struct gk20a *g, const struct nvgpu_device *dev,
|
|
bool enable);
|
|
|
|
int (*enable_devtype)(struct gk20a *g, u32 devtype, bool enable);
|
|
|
|
void (*fbpa_isr)(struct gk20a *g);
|
|
|
|
#ifdef CONFIG_NVGPU_LS_PMU
|
|
bool (*is_enabled)(struct gk20a *g, u32 unit);
|
|
#endif
|
|
|
|
bool (*is_intr1_pending)(struct gk20a *g, u32 unit, u32 mc_intr_1);
|
|
|
|
bool (*is_mmu_fault_pending)(struct gk20a *g);
|
|
|
|
bool (*is_intr_hub_pending)(struct gk20a *g, u32 mc_intr);
|
|
|
|
#ifdef CONFIG_NVGPU_NON_FUSA
|
|
void (*log_pending_intrs)(struct gk20a *g);
|
|
#endif
|
|
|
|
void (*fb_reset)(struct gk20a *g);
|
|
|
|
#ifdef CONFIG_NVGPU_DGPU
|
|
bool (*is_intr_nvlink_pending)(struct gk20a *g, u32 mc_intr);
|
|
#endif
|
|
|
|
/**
|
|
* @brief Reset HW engines.
|
|
*
|
|
* @param g [in] The GPU driver struct.
|
|
* @param devtype [in] Type of device.
|
|
*
|
|
* This function is invoked to reset the engines while initializing
|
|
* GR, CE and other engines during #nvgpu_finalize_poweron.
|
|
*
|
|
* Steps:
|
|
* - Compute reset mask for all engines of given devtype.
|
|
* - Disable given HW engines.
|
|
* - Acquire g->mc.enable_lock spinlock.
|
|
* - Read mc_device_enable_r register and clear the bits in read value
|
|
* corresponding to HW engines to be disabled.
|
|
* - Write mc_device_enable_r with the updated value.
|
|
* - Poll mc_device_enable_r to confirm register write success.
|
|
* - Release g->mc.enable_lock spinlock.
|
|
* - If GR engines are being reset, reset GPCs.
|
|
* - Enable the HW engines.
|
|
* - Acquire g->mc.enable_lock spinlock.
|
|
* - Read mc_device_enable_r register and set the bits in read value
|
|
* corresponding to HW engines to be enabled.
|
|
* - Write mc_device_enable_r with the updated value.
|
|
* - Poll mc_device_enable_r to confirm register write success.
|
|
* - Release g->mc.enable_lock spinlock.
|
|
*/
|
|
int (*reset_engines_all)(struct gk20a *g, u32 devtype);
|
|
void (*elpg_enable)(struct gk20a *g);
|
|
bool (*intr_get_unit_info)(struct gk20a *g, u32 unit);
|
|
|
|
/** @endcond DOXYGEN_SHOULD_SKIP_THIS */
|
|
};
|
|
|
|
#endif
|