Files
linux-nvgpu/drivers/gpu/nvgpu/common/ltc/ltc_gv11b.c
Antony Clince Alex b10960e7b7 gpu: nvgpu: Enable the reporting of ECC errors
Enable the reporting of ECC errors on hw modules
like gr, pmu and ltc. These errors will be notified
to the underlying safety service.

Jira NVGPU-1366

Change-Id: Ibf0f9761d30bcab31809f92aa2b4378360066385
Signed-off-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1955267
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: Raghuram Kothakota <rkothakota@nvidia.com>
Tested-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-by: Ankur Kishore <ankkishore@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2019-01-03 12:54:31 -08:00

258 lines
8.7 KiB
C

/*
* GV11B LTC
*
* Copyright (c) 2016-2018, 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.
*/
#include <nvgpu/io.h>
#include <nvgpu/gk20a.h>
#include <nvgpu/nvgpu_err.h>
#include "ltc_gp10b.h"
#include "ltc_gv11b.h"
#include <nvgpu/hw/gv11b/hw_ltc_gv11b.h>
#include <nvgpu/hw/gv11b/hw_mc_gv11b.h>
#include <nvgpu/hw/gv11b/hw_top_gv11b.h>
#include <nvgpu/hw/gv11b/hw_mc_gv11b.h>
#include <nvgpu/utils.h>
static void gv11b_ltc_report_ecc_error(struct gk20a *g, u32 ltc, u32 slice,
u32 err_type, u64 err_addr, u64 err_cnt)
{
int ret = 0;
u32 inst = 0U;
if (g->ops.ltc.err_ops.report_ecc_parity_err == NULL) {
return ;
}
if (slice < 256U) {
inst = (ltc << 8U) | slice;
} else {
nvgpu_err(g, "Invalid slice id=%u", slice);
return ;
}
ret = g->ops.ltc.err_ops.report_ecc_parity_err(g,
NVGPU_ERR_MODULE_LTC, inst, err_type, err_addr,
err_cnt);
if (ret != 0) {
nvgpu_err(g, "Failed to report LTC error: inst=%u, \
err_type=%u, err_addr=%llu, err_cnt=%llu",
inst, err_type, err_addr, err_cnt);
}
}
/*
* Sets the ZBC stencil for the passed index.
*/
void gv11b_ltc_set_zbc_stencil_entry(struct gk20a *g,
struct zbc_entry *stencil_val,
u32 index)
{
u32 real_index = index + GK20A_STARTOF_ZBC_TABLE;
nvgpu_writel_check(g, ltc_ltcs_ltss_dstg_zbc_index_r(),
ltc_ltcs_ltss_dstg_zbc_index_address_f(real_index));
nvgpu_writel_check(g,
ltc_ltcs_ltss_dstg_zbc_stencil_clear_value_r(),
stencil_val->depth);
}
void gv11b_ltc_init_fs_state(struct gk20a *g)
{
struct gr_gk20a *gr = &g->gr;
u32 ltc_intr;
u32 reg;
nvgpu_log_info(g, "initialize gv11b l2");
g->max_ltc_count = gk20a_readl(g, top_num_ltcs_r());
g->ltc_count = g->ops.priv_ring.enum_ltc(g);
nvgpu_log_info(g, "%u ltcs out of %u", g->ltc_count, g->max_ltc_count);
reg = gk20a_readl(g, ltc_ltcs_ltss_cbc_param_r());
gr->slices_per_ltc = ltc_ltcs_ltss_cbc_param_slices_per_ltc_v(reg);;
gr->cacheline_size =
U32(512) << ltc_ltcs_ltss_cbc_param_cache_line_size_v(reg);
/* Disable LTC interrupts */
reg = gk20a_readl(g, ltc_ltcs_ltss_intr_r());
reg &= ~ltc_ltcs_ltss_intr_en_evicted_cb_m();
reg &= ~ltc_ltcs_ltss_intr_en_illegal_compstat_access_m();
nvgpu_writel_check(g, ltc_ltcs_ltss_intr_r(), reg);
if (g->ops.ltc.intr_en_illegal_compstat != NULL) {
g->ops.ltc.intr_en_illegal_compstat(g,
g->ltc_intr_en_illegal_compstat);
}
/* Enable ECC interrupts */
ltc_intr = gk20a_readl(g, ltc_ltcs_ltss_intr_r());
ltc_intr |= ltc_ltcs_ltss_intr_en_ecc_sec_error_enabled_f() |
ltc_ltcs_ltss_intr_en_ecc_ded_error_enabled_f();
nvgpu_writel_check(g, ltc_ltcs_ltss_intr_r(),
ltc_intr);
}
void gv11b_ltc_intr_en_illegal_compstat(struct gk20a *g, bool enable)
{
u32 val;
/* disble/enble illegal_compstat interrupt */
val = gk20a_readl(g, ltc_ltcs_ltss_intr_r());
if (enable) {
val = set_field(val,
ltc_ltcs_ltss_intr_en_illegal_compstat_m(),
ltc_ltcs_ltss_intr_en_illegal_compstat_enabled_f());
} else {
val = set_field(val,
ltc_ltcs_ltss_intr_en_illegal_compstat_m(),
ltc_ltcs_ltss_intr_en_illegal_compstat_disabled_f());
}
gk20a_writel(g, ltc_ltcs_ltss_intr_r(), val);
}
void gv11b_ltc_lts_isr(struct gk20a *g, unsigned int ltc, unsigned int slice)
{
u32 offset;
u32 ltc_intr3;
u32 ecc_status, ecc_addr, corrected_cnt, uncorrected_cnt;
u32 corrected_delta, uncorrected_delta;
u32 corrected_overflow, uncorrected_overflow;
u32 ltc_stride = nvgpu_get_litter_value(g, GPU_LIT_LTC_STRIDE);
u32 lts_stride = nvgpu_get_litter_value(g, GPU_LIT_LTS_STRIDE);
offset = ltc_stride * ltc + lts_stride * slice;
ltc_intr3 = gk20a_readl(g, ltc_ltc0_lts0_intr3_r() +
offset);
/* Detect and handle ECC PARITY errors */
if ((ltc_intr3 &
(ltc_ltcs_ltss_intr3_ecc_uncorrected_m() |
ltc_ltcs_ltss_intr3_ecc_corrected_m())) != 0U) {
ecc_status = gk20a_readl(g,
ltc_ltc0_lts0_l2_cache_ecc_status_r() +
offset);
ecc_addr = gk20a_readl(g,
ltc_ltc0_lts0_l2_cache_ecc_address_r() +
offset);
corrected_cnt = gk20a_readl(g,
ltc_ltc0_lts0_l2_cache_ecc_corrected_err_count_r() + offset);
uncorrected_cnt = gk20a_readl(g,
ltc_ltc0_lts0_l2_cache_ecc_uncorrected_err_count_r() + offset);
corrected_delta =
ltc_ltc0_lts0_l2_cache_ecc_corrected_err_count_total_v(corrected_cnt);
uncorrected_delta =
ltc_ltc0_lts0_l2_cache_ecc_uncorrected_err_count_total_v(uncorrected_cnt);
corrected_overflow = ecc_status &
ltc_ltc0_lts0_l2_cache_ecc_status_corrected_err_total_counter_overflow_m();
uncorrected_overflow = ecc_status &
ltc_ltc0_lts0_l2_cache_ecc_status_uncorrected_err_total_counter_overflow_m();
/* clear the interrupt */
if ((corrected_delta > 0U) || (corrected_overflow != 0U)) {
nvgpu_writel_check(g,
ltc_ltc0_lts0_l2_cache_ecc_corrected_err_count_r() + offset, 0);
}
if ((uncorrected_delta > 0U) || (uncorrected_overflow !=0U)) {
nvgpu_writel_check(g,
ltc_ltc0_lts0_l2_cache_ecc_uncorrected_err_count_r() + offset, 0);
}
nvgpu_writel_check(g,
ltc_ltc0_lts0_l2_cache_ecc_status_r() + offset,
ltc_ltc0_lts0_l2_cache_ecc_status_reset_task_f());
/* update counters per slice */
if (corrected_overflow != 0U) {
corrected_delta += BIT32(ltc_ltc0_lts0_l2_cache_ecc_corrected_err_count_total_s());
}
if (uncorrected_overflow != 0U) {
uncorrected_delta += BIT32(ltc_ltc0_lts0_l2_cache_ecc_uncorrected_err_count_total_s());
}
g->ecc.ltc.ecc_sec_count[ltc][slice].counter += corrected_delta;
g->ecc.ltc.ecc_ded_count[ltc][slice].counter += uncorrected_delta;
nvgpu_log(g, gpu_dbg_intr,
"ltc:%d lts: %d cache ecc interrupt intr: 0x%x", ltc, slice, ltc_intr3);
if ((ecc_status & ltc_ltc0_lts0_l2_cache_ecc_status_corrected_err_rstg_m()) != 0U) {
gv11b_ltc_report_ecc_error(g, ltc, slice,
LTC_CACHE_RSTG_ECC_CORRECTED, ecc_addr,
g->ecc.ltc.ecc_sec_count[ltc][slice].counter);
nvgpu_log(g, gpu_dbg_intr, "rstg ecc error corrected");
}
if ((ecc_status & ltc_ltc0_lts0_l2_cache_ecc_status_uncorrected_err_rstg_m()) != 0U) {
gv11b_ltc_report_ecc_error(g, ltc, slice,
LTC_CACHE_RSTG_ECC_UNCORRECTED, ecc_addr,
g->ecc.ltc.ecc_ded_count[ltc][slice].counter);
nvgpu_log(g, gpu_dbg_intr, "rstg ecc error uncorrected");
}
if ((ecc_status & ltc_ltc0_lts0_l2_cache_ecc_status_corrected_err_tstg_m()) != 0U) {
gv11b_ltc_report_ecc_error(g, ltc, slice,
LTC_CACHE_TSTG_ECC_CORRECTED, ecc_addr,
g->ecc.ltc.ecc_sec_count[ltc][slice].counter);
nvgpu_log(g, gpu_dbg_intr, "tstg ecc error corrected");
}
if ((ecc_status & ltc_ltc0_lts0_l2_cache_ecc_status_uncorrected_err_tstg_m()) != 0U) {
gv11b_ltc_report_ecc_error(g, ltc, slice,
LTC_CACHE_TSTG_ECC_UNCORRECTED, ecc_addr,
g->ecc.ltc.ecc_ded_count[ltc][slice].counter);
nvgpu_log(g, gpu_dbg_intr, "tstg ecc error uncorrected");
}
if ((ecc_status & ltc_ltc0_lts0_l2_cache_ecc_status_corrected_err_dstg_m()) != 0U) {
gv11b_ltc_report_ecc_error(g, ltc, slice,
LTC_CACHE_DSTG_ECC_CORRECTED, ecc_addr,
g->ecc.ltc.ecc_sec_count[ltc][slice].counter);
nvgpu_log(g, gpu_dbg_intr, "dstg ecc error corrected");
}
if ((ecc_status & ltc_ltc0_lts0_l2_cache_ecc_status_uncorrected_err_dstg_m()) != 0U) {
gv11b_ltc_report_ecc_error(g, ltc, slice,
LTC_CACHE_DSTG_ECC_UNCORRECTED, ecc_addr,
g->ecc.ltc.ecc_ded_count[ltc][slice].counter);
nvgpu_log(g, gpu_dbg_intr, "dstg ecc error uncorrected");
}
if ((corrected_overflow != 0U) || (uncorrected_overflow != 0U)) {
nvgpu_info(g, "ecc counter overflow!");
}
nvgpu_log(g, gpu_dbg_intr,
"ecc error address: 0x%x", ecc_addr);
}
gp10b_ltc_lts_isr(g, ltc, slice);
}
void gv11b_ltc_isr(struct gk20a *g, unsigned int ltc)
{
unsigned int slice;
for (slice = 0U; slice < g->gr.slices_per_ltc; slice++) {
gv11b_ltc_lts_isr(g, ltc, slice);
}
}