diff options
Diffstat (limited to 'src/import/chips/p9/procedures/hwp/memory/lib/phy')
| -rw-r--r-- | src/import/chips/p9/procedures/hwp/memory/lib/phy/ddr_phy.C | 11 | ||||
| -rw-r--r-- | src/import/chips/p9/procedures/hwp/memory/lib/phy/dp16.C | 42 |
2 files changed, 48 insertions, 5 deletions
diff --git a/src/import/chips/p9/procedures/hwp/memory/lib/phy/ddr_phy.C b/src/import/chips/p9/procedures/hwp/memory/lib/phy/ddr_phy.C index 86a8621fa..e1e63fec5 100644 --- a/src/import/chips/p9/procedures/hwp/memory/lib/phy/ddr_phy.C +++ b/src/import/chips/p9/procedures/hwp/memory/lib/phy/ddr_phy.C @@ -521,6 +521,11 @@ fapi2::ReturnCode process_initial_cal_errors( const fapi2::Target<TARGET_TYPE_DI uint64_t l_rank_pairs = 0; uint8_t cal_abort_on_error = 0; + // This boolean tells the code whether we took a training fail or a scom fail reading the status registers + // It starts as false, given that we need to read out the registers + // When we start checking all of the values of the status registers, it gets set to true + bool l_check_firs = false; + const auto& l_mca = mss::find_target<fapi2::TARGET_TYPE_MCA>(i_target); fapi2::buffer<uint64_t> l_err_data; @@ -550,6 +555,9 @@ fapi2::ReturnCode process_initial_cal_errors( const fapi2::Target<TARGET_TYPE_DI } // Error information from other registers is gathered in the FFDC from the XML + // From here on out, check the FIRs + // Using this boolean to avoid having to check the FIR's after each assert below + l_check_firs = true; // So we can do a few things here. If we're aborting on the first calibration error, // we only expect to have one error bit set. If we ran all the calibrations, we can @@ -692,7 +700,8 @@ fapi_try_exit: (fapi2::current_err == fapi2::FAPI2_RC_SUCCESS ? "success" : "errors reported"), mss::c_str(l_mca)); - return fapi2::current_err; + // Checks the FIR's, if need be + return mss::check::fir_or_pll_fail( i_target, fapi2::current_err, l_check_firs); } /// diff --git a/src/import/chips/p9/procedures/hwp/memory/lib/phy/dp16.C b/src/import/chips/p9/procedures/hwp/memory/lib/phy/dp16.C index 0e346881a..129c37515 100644 --- a/src/import/chips/p9/procedures/hwp/memory/lib/phy/dp16.C +++ b/src/import/chips/p9/procedures/hwp/memory/lib/phy/dp16.C @@ -52,6 +52,7 @@ #include <generic/memory/lib/utils/c_str.H> #include <lib/workarounds/dp16_workarounds.H> +#include <lib/fir/check.H> #include <generic/memory/lib/utils/mss_math.H> using fapi2::TARGET_TYPE_MCS; @@ -3260,6 +3261,22 @@ fapi_try_exit: /// fapi2::ReturnCode record_bad_bits( const fapi2::Target<fapi2::TARGET_TYPE_MCA>& i_target ) { + // If we have a FIR set that could have caused our training fail, then skip checking bad bits in FW + // PRD will handle the FIR and retrigger the procedure +#ifdef __HOSTBOOT_MODULE + bool l_fir_error = false; + FAPI_TRY(mss::check::bad_fir_bits(i_target, l_fir_error), "%s took an error while checking FIR's", + mss::c_str(i_target)); + + // Exit if we took a FIR error - PRD will handle bad bits + if(l_fir_error) + { + FAPI_INF("%s has FIR's set, exiting to let PRD handle it", mss::c_str(i_target)); + return fapi2::FAPI2_RC_SUCCESS; + } + +#endif + for( const auto& d : mss::find_targets<fapi2::TARGET_TYPE_DIMM>(i_target) ) { uint8_t l_data[MAX_RANK_PER_DIMM][BAD_DQ_BYTE_COUNT] = {}; @@ -3367,11 +3384,17 @@ fapi2::ReturnCode process_rdvref_cal_errors( const fapi2::Target<fapi2::TARGET_T size_t l_index = 0; std::vector<fapi2::buffer<uint64_t>> l_data; + // Boolean to keep track of if a fail was calibration related, or scom related + bool l_cal_fail = false; + // Suck all the cal error bits out ... FAPI_TRY( mss::scom_suckah(l_mca, TT::RD_VREF_CAL_ERROR_REG, l_data) ); FAPI_INF("%s Processing RD_VREF_CAL_ERROR", mss::c_str(i_target)); + // From here on out, the FIR's are all cal fails + l_cal_fail = true; + for (const auto& v : l_data) { // They should all be 0's. If they're not, we have a problem. @@ -3383,14 +3406,17 @@ fapi2::ReturnCode process_rdvref_cal_errors( const fapi2::Target<fapi2::TARGET_T .set_VALUE(v), "DP16 failed read vref calibration on %s. register 0x%016lx value 0x%016lx", mss::c_str(l_mca), TT::RD_VREF_CAL_ERROR_REG[l_index], v); + ++l_index; } - FAPI_INF("RD_VREF_CAL_ERROR complete"); + FAPI_INF("%s RD_VREF_CAL_ERROR complete", mss::c_str(i_target)); return fapi2::FAPI2_RC_SUCCESS; fapi_try_exit: - return fapi2::current_err; + + // If the FIR's are cal fails, then check to see if FIRs or PLL fails were the cause + return mss::check::fir_or_pll_fail( i_target, fapi2::current_err, l_cal_fail); } /// @@ -3412,10 +3438,16 @@ fapi2::ReturnCode process_wrvref_cal_errors( const fapi2::Target<fapi2::TARGET_T std::vector<std::pair<fapi2::buffer<uint64_t>, fapi2::buffer<uint64_t>>> l_data; std::vector<std::pair<fapi2::buffer<uint64_t>, fapi2::buffer<uint64_t>>> l_mask; + // Boolean to keep track of if a fail was calibration related, or scom related + bool l_cal_fail = false; + // Suck all the cal error bits out ... FAPI_TRY( mss::scom_suckah(l_mca, TT::WR_VREF_ERROR_REG, l_data) ); FAPI_TRY( mss::scom_suckah(l_mca, TT::WR_VREF_ERROR_MASK_REG, l_mask) ); + // From here on out, the FIR's are all cal fails + l_cal_fail = true; + // Loop through both data and mask { // Note: ideally these would be cbegin/cend, but HB doesn't support constant iterators for vectors @@ -3480,11 +3512,13 @@ fapi2::ReturnCode process_wrvref_cal_errors( const fapi2::Target<fapi2::TARGET_T } } - FAPI_INF("WRVREF_CAL_ERROR complete"); + FAPI_INF("%s WRVREF_CAL_ERROR complete", mss::c_str(i_target)); return fapi2::FAPI2_RC_SUCCESS; fapi_try_exit: - return fapi2::current_err; + + // If the FIR's are cal fails, then check to see if FIR's were the cause + return mss::check::fir_or_pll_fail( i_target, fapi2::current_err, l_cal_fail); } /// |
