git: 9ff3e1cd3179 - stable/15 - e1000: Report 82575 memory ECC errors

The branch stable/15 has been updated by kbowling:

URL: https://cgit.FreeBSD.org/src/commit/?id=9ff3e1cd31792202c0e7c7fdaf877e1fc6c54269

commit 9ff3e1cd31792202c0e7c7fdaf877e1fc6c54269
Author:     Kevin Bowling <kbowling@FreeBSD.org>
AuthorDate: 2026-08-16 05:40:44 +0000
Commit:     Kevin Bowling <kbowling@FreeBSD.org>
CommitDate: 2026-08-30 01:28:02 +0000

    e1000: Report 82575 memory ECC errors
    
    82575 exposes clear-on-read, saturating counters for corrected and
    uncorrected errors in the packet buffer and the receive and transmit
    descriptor handlers.
    
    Sample all three status registers together from the regular hardware
    statistics update.  When an unrecoverable event interrupts first, count
    the values captured by the interrupt filter so the clear-on-read status
    is not lost before the admin task handles it.
    
    Expose packet buffer and descriptor handler counters under the existing
    memory_errors sysctl node.
    
    Hardware validation used an 82575EB revision 2 and the documented PBEEI,
    RDHEEI, and TDHEEI injectors.  Correctable and uncorrectable TX/RX
    packet-buffer errors and receive/transmit descriptor-handler errors
    advanced the corresponding counters.  The controls and accounting
    survived repeated recovery resets and an ordinary interface down/up.
    
    Sponsored by:   BBOX.io
    
    (cherry picked from commit 7accd803e28cf6051cac69e614eec6f24d5c77cc)
---
 sys/dev/e1000/if_em.c | 48 +++++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 45 insertions(+), 3 deletions(-)

diff --git a/sys/dev/e1000/if_em.c b/sys/dev/e1000/if_em.c
index c4c036ff17f7..10a2e97a8c02 100644
--- a/sys/dev/e1000/if_em.c
+++ b/sys/dev/e1000/if_em.c
@@ -2035,6 +2035,26 @@ em_fatal_error_intr_mask(struct e1000_softc *sc)
 	return (em_memory_error_intr_mask(&sc->hw));
 }
 
+static void
+em_update_82575_ecc_stats(struct e1000_softc *sc, u32 pbeccsts,
+    u32 rdhests, u32 tdhests)
+{
+
+	sc->corrected_error_packet_buffer_count +=
+	    pbeccsts & E1000_ECC_82575_CORR_CNT_MASK;
+	sc->uncorrected_error_packet_buffer_count +=
+	    (pbeccsts & E1000_ECC_82575_UNCORR_CNT_MASK) >>
+	    E1000_ECC_82575_UNCORR_CNT_SHIFT;
+	sc->corrected_error_dma_count +=
+	    (rdhests & E1000_ECC_82575_CORR_CNT_MASK) +
+	    (tdhests & E1000_ECC_82575_CORR_CNT_MASK);
+	sc->uncorrected_error_dma_count +=
+	    ((rdhests & E1000_ECC_82575_UNCORR_CNT_MASK) >>
+	    E1000_ECC_82575_UNCORR_CNT_SHIFT) +
+	    ((tdhests & E1000_ECC_82575_UNCORR_CNT_MASK) >>
+	    E1000_ECC_82575_UNCORR_CNT_SHIFT);
+}
+
 static void
 em_update_82576_ecc_counter(struct e1000_softc *sc, u32 reg,
     u64 *corrected, u64 *uncorrected)
@@ -2267,6 +2287,8 @@ em_handle_fatal_error_admin(struct e1000_softc *sc)
 		    "PBECCSTS %#x; requesting reset\n",
 		    sc->fatal_error_pbeccsts);
 	} else if (em_has_82575_memory_errors(&sc->hw)) {
+		em_update_82575_ecc_stats(sc, sc->fatal_error_pbeccsts,
+		    sc->fatal_error_dma_rx, sc->fatal_error_dma_tx);
 		device_printf(sc->dev,
 		    "unrecoverable internal memory ECC error: ICR %#x, "
 		    "PBECCSTS %#x, RDHESTS %#x, TDHESTS %#x; "
@@ -5733,6 +5755,11 @@ em_update_stats_counters(struct e1000_softc *sc)
 	if (em_has_pch_ecc(&sc->hw))
 		em_update_pch_ecc_stats(sc,
 		    E1000_READ_REG(&sc->hw, E1000_PBECCSTS));
+	else if (em_has_82575_memory_errors(&sc->hw))
+		em_update_82575_ecc_stats(sc,
+		    E1000_READ_REG(&sc->hw, E1000_PBECCSTS_82575),
+		    E1000_READ_REG(&sc->hw, E1000_RDHESTS_82575),
+		    E1000_READ_REG(&sc->hw, E1000_TDHESTS_82575));
 	else if (em_has_82576_memory_errors(&sc->hw))
 		em_update_82576_ecc_stats(sc);
 	else if (em_has_i350_memory_errors(&sc->hw))
@@ -6042,9 +6069,7 @@ em_add_hw_stats(struct e1000_softc *sc)
 		SYSCTL_ADD_UQUAD(ctx, eee_list, OID_AUTO, "rx_lpi_count",
 		    CTLFLAG_RD, &stats->rlpic, "RX LPI event count");
 	}
-	if (em_has_pch_ecc(&sc->hw) ||
-	    em_has_82576_memory_errors(&sc->hw) ||
-	    em_has_i210_i350_memory_errors(&sc->hw)) {
+	if (em_has_memory_errors(&sc->hw)) {
 		struct sysctl_oid *memerr_node;
 		struct sysctl_oid_list *memerr_list;
 
@@ -6065,6 +6090,23 @@ em_add_hw_stats(struct e1000_softc *sc)
 			    "uncorrected_packet_buffer", CTLFLAG_RD,
 			    &sc->uncorrected_error_packet_buffer_count,
 			    "Uncorrected packet-buffer ECC errors");
+		} else if (em_has_82575_memory_errors(&sc->hw)) {
+			SYSCTL_ADD_UQUAD(ctx, memerr_list, OID_AUTO,
+			    "corrected_packet_buffer", CTLFLAG_RD,
+			    &sc->corrected_error_packet_buffer_count,
+			    "Corrected packet-buffer ECC errors");
+			SYSCTL_ADD_UQUAD(ctx, memerr_list, OID_AUTO,
+			    "uncorrected_packet_buffer", CTLFLAG_RD,
+			    &sc->uncorrected_error_packet_buffer_count,
+			    "Uncorrected packet-buffer ECC errors");
+			SYSCTL_ADD_UQUAD(ctx, memerr_list, OID_AUTO,
+			    "corrected_descriptor_handler", CTLFLAG_RD,
+			    &sc->corrected_error_dma_count,
+			    "Corrected descriptor-handler ECC errors");
+			SYSCTL_ADD_UQUAD(ctx, memerr_list, OID_AUTO,
+			    "uncorrected_descriptor_handler", CTLFLAG_RD,
+			    &sc->uncorrected_error_dma_count,
+			    "Uncorrected descriptor-handler ECC errors");
 		} else if (em_has_82576_memory_errors(&sc->hw)) {
 			SYSCTL_ADD_UQUAD(ctx, memerr_list, OID_AUTO,
 			    "fatal_unknown", CTLFLAG_RD,