Attention is currently required from: Furquan Shaikh, Tim Wawrzynczak, Paul Menzel. Sugnan Prabhu S has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/56750 )
Change subject: vc/google/chromeos: Add support for new SAR tables revisions ......................................................................
Patch Set 17:
(2 comments)
File src/include/sar.h:
https://review.coreboot.org/c/coreboot/+/56750/comment/bbe04b66_cb174827 PS17, Line 21: struct wifi_sar_delta_table { : uint8_t version; : union { : struct { : uint8_t power_max_2400mhz; : uint8_t power_chain_a_2400mhz; : uint8_t power_chain_b_2400mhz; : uint8_t power_max_5200mhz; : uint8_t power_chain_a_5200mhz; : uint8_t power_chain_b_5200mhz; : } __packed group_rev0[SAR_NUM_WGDS_GROUPS]; : struct { : uint8_t power_max_2400mhz; : uint8_t power_chain_a_2400mhz; : uint8_t power_chain_b_2400mhz; : uint8_t power_max_5200mhz; : uint8_t power_chain_a_5200mhz; : uint8_t power_chain_b_5200mhz; : uint8_t power_max_6000mhz; : uint8_t power_chain_a_6000mhz; : uint8_t power_chain_b_6000mhz; : } __packed group_rev1[SAR_NUM_WGDS_GROUPS]; : } __packed; : } __packed;
I think we should hide the differences of different revisions on the SAR file generation side so tha […]
This think this will be better approach. I will update all the current implementation as per the suggestion.
File src/vendorcode/google/chromeos/sar.c:
https://review.coreboot.org/c/coreboot/+/56750/comment/b6bc7e9e_cb53301c PS17, Line 138: if (CONFIG(USE_SAR)) { : sar_limits->sar_table->version = 0; : sar_limits->sar_table->sar_enable = CONFIG(SAR_ENABLE); : sar_limits->sar_table->dsar_enable = CONFIG(DSAR_ENABLE); : sar_limits->sar_table->dsar_set_num = CONFIG_DSAR_SET_NUM; : }
I like the fact that you are trying to fix the problems with the original representation of SAR tabl […]
I will update the logic to remove USE_SAR_V2. Only possible SAR file size with the older implementation is 119 and 81. As of now, none of the revision combination leads to this size of binary, I will recheck once I address the review comments. I think we can also have some fallback mechanism to use the other version incase if the decoding fails with revision 0.