Hi Ben,

On 29/08/2025 16:55, Ben Horgan wrote:

On 8/22/25 16:30, James Morse wrote:

quoted

Reading a monitor involves configuring what you want to monitor, and
reading the value. Components made up of multiple MSC may need values
from each MSC. MSCs may take time to configure, returning 'not ready'.
The maximum 'not ready' time should have been provided by firmware.

Add mpam_msmon_read() to hide all this. If (one of) the MSC returns
not ready, then wait the full timeout value before trying again.

quoted

diff --git a/drivers/resctrl/mpam_devices.c b/drivers/resctrl/mpam_devices.c
index e7e00c632512..9ce771aaf671 100644
--- a/drivers/resctrl/mpam_devices.c
+++ b/drivers/resctrl/mpam_devices.c

@@ -973,6 +973,228 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)

quoted

+static void gen_msmon_ctl_flt_vals(struct mon_read *m, u32 *ctl_val,
+				   u32 *flt_val)
+{
+	struct mon_cfg *ctx = m->ctx;
+
+	switch (m->type) {
+	case mpam_feat_msmon_csu:
+		*ctl_val = MSMON_CFG_CSU_CTL_TYPE_CSU;
+		break;
+	case mpam_feat_msmon_mbwu:
+		*ctl_val = MSMON_CFG_MBWU_CTL_TYPE_MBWU;
+		break;
+	default:
+		return;
+	}
+
+	/*
+	 * For CSU counters its implementation-defined what happens when not
+	 * filtering by partid.
+	 */
+	*ctl_val |= MSMON_CFG_x_CTL_MATCH_PARTID;
+
+	*flt_val = FIELD_PREP(MSMON_CFG_MBWU_FLT_PARTID, ctx->partid);
+	if (m->ctx->match_pmg) {
+		*ctl_val |= MSMON_CFG_x_CTL_MATCH_PMG;
+		*flt_val |= FIELD_PREP(MSMON_CFG_MBWU_FLT_PMG, ctx->pmg);
+	}

As we are using MSMON_CFG_MBWU_FLT_{PMG,PARTID} for both CSU and MBWU
how about changing to MSMON_CFG_x_FLT_{PMG,PARTID}?

Sure,

quoted

+
+	if (mpam_has_feature(mpam_feat_msmon_mbwu_rwbw, &m->ris->props))
+		*flt_val |= FIELD_PREP(MSMON_CFG_MBWU_FLT_RWBW, ctx->opts);

This needs to be conditional on the type of the monitor being
configured. There is an XCL bit here for CSU monitors.

Fixed ... that wasn't there last time I looked!
I may as well wire that up too...


Thanks,

James