Preload bad pages occur during the sw_init stage, so
the preload-related code in RAS module must be moved
to sw_init for proper initialization.
v2:
Rename functions and variables.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rework fw eeprom handling for new management interface:
- Support new eeprom management interfaces.
- Use ras_mp1 common interfaces to retrieve bad page information
- Keep a local backup to reduce the interaction frequency with MP1.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Switch to use new eeprom management interfaces.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
- Refactor EEPROM parameter config to support multiple ras EEPROMs.
- Convert bad page count to record count as EEPROM threshold.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Support retrieving bad page info from mp1_v15_0.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Ras mp1 introduces common interfaces to manage mp1 data
obtained from different ASICs.
v2:
Add mutex to ensure complete data transactions with MP1.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make space at the end of psp->cmd_buf_bo usable for extra data returned
from PSP commands.
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Mukul Joshi <mukul.joshi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add ext_umc_mask for the upper half of
adev->umc.active_mask and fill both fields
when building RAS TA init message.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add ext_umc_mask for the upper half of
adev->umc.active_mask and fill both fields
when building RAS TA init message.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
amdgpu_umc.active_mask is widened to u64 to
match 64-bit harvest masks and avoid LP32
unsigned long width limits.
Fix UMC harvest bitmask (drop le16 on u8, use 1ULL shift).
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Register psp_15_0_8_ta.bin with MODULE_FIRMWARE and
load TA for psp v15_0_8.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add psp_init_rl_microcode() to request amdgpu/%s_rl.bin and populate
psp->rl using the existing v1.0 PSP firmware header parsing.
Wire it up for psp v15_0_8.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add parse_psp_v1_bin_descriptor to populate struct psp_bin_desc
from psp_firmware_header_v1_0, shared by ASD, TOC and CAP
standalone firmware loads.
Use psp-> fields directly instead of adev->psp.
v2: code rebase
v3: Rename v1 header helper to parse_psp_v1_bin_descriptor,
have it return an error code with explicit NULL checks.
Rename sos v2 helper to parse_psp_v2_bin_descriptor for symmetry.
Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Support poison consumption handling for A + A sriov guest.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some ASICs have different hw ras block IDs.
In order to be compatible with these ASICs,
it is necessary to map these hw RAS block IDs
to unified RAS block IDs.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add ras_psp v15_0 for A + A.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add ras ip configuration for A + A.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add new ras error type so that ACA parser
can parse ras data from mce.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add aca parser v5_0.
V2:
Support new mappings for hwip and hwid.
V3:
Change aca ip version.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Improve function interface to record different types of data.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Support separate logging of gfx aid and xcd errors.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Introduce new fields for BOOT CPER encoding, only important fields are
introduced here due to the cache size limit.
Signed-off-by: Xiang Liu <xiang.liu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Introduce MCE type ras log event.
Parse error type of MCE and encode to corresponding CPER.
Signed-off-by: Xiang Liu <xiang.liu@amd.com>
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Register mce notifier to obtain ras error info for A + A.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Signed-off-by: Xiang Liu <xiang.liu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
strcpy() has been deprecated because it performs no bounds checking
on the destination buffer, which can lead to buffer overflows. Replace
it with the safer strscpy(). No functional changes.
Signed-off-by: Amin Gattout <amin.gattout@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove the actual is_idle member, since it's never read/called and
the previous patches have remove all the assignments.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The is_idle method on amd_ip_funcs is never called.
Remove the implementations that actually do something.
In most cases the same code has been duplicated in
the wait_for_idle method implementations that is still used.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The is_idle method on amd_ip_funcs is never called.
Remove the trivial cases where either:
a) It's just a return true function that can be deleted
together with the .is_init =
b) The function is called somewhere else, so we just remove
the assignment.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
