diff options
| author | Hawking Zhang <Hawking.Zhang@amd.com> | 2025-01-22 19:34:33 +0800 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2025-02-12 21:02:59 -0500 |
| commit | 16b85a0942c0b0f1611bcaa42cc98f020e34b1cf (patch) | |
| tree | d7f35626c9938478c28b84fea369176192559412 /drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | |
| parent | c003b5ccaf625c4e8077a0e7a8a1d9e6e403d603 (diff) | |
| download | lwn-16b85a0942c0b0f1611bcaa42cc98f020e34b1cf.tar.gz lwn-16b85a0942c0b0f1611bcaa42cc98f020e34b1cf.zip | |
drm/amdgpu: Update usage for bad page threshold
The driver's behavior varies based on
the configuration of amdgpu_bad_page_threshold setting
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 40 |
1 files changed, 20 insertions, 20 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index f0924aa3f4e4..90394f89aba6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -3080,31 +3080,29 @@ static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev, struct amdgpu_ras *con = amdgpu_ras_get_context(adev); /* - * Justification of value bad_page_cnt_threshold in ras structure - * - * Generally, 0 <= amdgpu_bad_page_threshold <= max record length - * in eeprom or amdgpu_bad_page_threshold == -2, introduce two - * scenarios accordingly. - * - * Bad page retirement enablement: - * - If amdgpu_bad_page_threshold = -2, - * bad_page_cnt_threshold = typical value by formula. - * - * - When the value from user is 0 < amdgpu_bad_page_threshold < - * max record length in eeprom, use it directly. - * - * Bad page retirement disablement: - * - If amdgpu_bad_page_threshold = 0, bad page retirement - * functionality is disabled, and bad_page_cnt_threshold will - * take no effect. + * amdgpu_bad_page_threshold is used to config + * the threshold for the number of bad pages. + * -1: Threshold is set to default value + * Driver will issue a warning message when threshold is reached + * and continue runtime services. + * 0: Disable bad page retirement + * Driver will not retire bad pages + * which is intended for debugging purpose. + * -2: Threshold is determined by a formula + * that assumes 1 bad page per 100M of local memory. + * Driver will continue runtime services when threhold is reached. + * 0 < threshold < max number of bad page records in EEPROM, + * A user-defined threshold is set + * Driver will halt runtime services when this custom threshold is reached. */ - - if (amdgpu_bad_page_threshold < 0) { + if (amdgpu_bad_page_threshold == -2) { u64 val = adev->gmc.mc_vram_size; do_div(val, RAS_BAD_PAGE_COVER); con->bad_page_cnt_threshold = min(lower_32_bits(val), max_count); + } else if (amdgpu_bad_page_threshold == -1) { + con->bad_page_cnt_threshold = ((con->reserved_pages_in_bytes) >> 21) << 4; } else { con->bad_page_cnt_threshold = min_t(int, max_count, amdgpu_bad_page_threshold); @@ -3848,8 +3846,10 @@ static void amdgpu_ras_init_reserved_vram_size(struct amdgpu_device *adev) case IP_VERSION(13, 0, 2): case IP_VERSION(13, 0, 6): case IP_VERSION(13, 0, 12): + con->reserved_pages_in_bytes = AMDGPU_RAS_RESERVED_VRAM_SIZE_DEFAULT; + break; case IP_VERSION(13, 0, 14): - con->reserved_pages_in_bytes = AMDGPU_RAS_RESERVED_VRAM_SIZE; + con->reserved_pages_in_bytes = (AMDGPU_RAS_RESERVED_VRAM_SIZE_DEFAULT << 1); break; default: break; |