Missing decode definitions for Product ID 0x037: PMC rev 5 / PMC_TRACE rev 1 / PMC_RST rev 1

[scoommy]

Summary

I am debugging repeated hard freezes on an Intel 13th Gen Core desktop platform using a Z690 / Intel 600-series PCH. After one freeze, Linux reported an ACPI BERT hardware error record. The Boot Error Region contains three Intel Crash Log records with Product ID 0x037:

  #   Record Type  Rev.  Product  Size  Skt  Checksum  Die
----- ------------ ----- -------- ----- ---- --------- ----
 0-0  PMC              5 0x037     2508    0 Valid
 1-0  PMC_TRACE        1 0x037      512    0
 2-0  PMC_RST          1 0x037       40    0

iclg triage reports:

CRASHLOG_REASON.PMC.30010H

iclg decode only decodes the common record headers; there does not appear to be public collateral for Product ID 0x037, PMC revision 5.

Environment
iclg / intel-crashlog: v1.1.0
Linux/TrueNAS: TrueNAS 26.0.0-BETA.3
Kernel: 6.18.42-production+truenas
ACPI BERT record obtained from Linux sysfs after a hard freeze and manual reboot

The Linux file /sys/firmware/acpi/tables/data/BERT contained the raw Boot Error Region rather than a complete BERT table/file wrapper. To make the saved copy readable by iclg, I prepended the four ASCII bytes BERR, matching Berr::from_bert_file() in the library source. iclg info, triage, decode, and unpack then parsed it successfully.

Hardware / platform

• CPU: Intel Core i7-13700TE (13th Gen Core), with Intel UHD Graphics 770 integrated graphics
• Memory: 32 GB DDR4 (approximately 31 GiB reported by TrueNAS), non-ECC
• Motherboard/platform: LGA1700 Z690 platform, Intel 600-series PCH; board used for this build is a BKHD Z690-MATX
• Onboard networking: 4 × Intel I226-V 2.5GbE controllers (8086:125c) at PCI addresses 02:00.0, 03:00.0, 04:00.0, and 05:00.0
• The four onboard I226-V controllers were disabled in BIOS during the freeze isolation
• Active add-in NIC during the captured failure: TP-Link TX201, Realtek RTL8125 rev 04 at 01:00.0, using the Linux r8169 driver
• Boot device: WDC PC SN530 256 GB NVMe
• Main storage: 4 × 3 TB SATA HDDs in a RAIDZ1 pool, plus a separate 500 GB SATA disk used for applications/VM storage
• No discrete GPU is installed; workloads use the Intel integrated GPU
• Relevant platform configuration at the time of the captured freeze: CPU C-states disabled, SATA ALPM disabled, PCIe ASPM enabled

The onboard Intel I226-V controllers were initially suspected because they had shown PCIe/AER and transmit-stall symptoms. However, a hard freeze also occurred with all four I226-V controllers disabled and the Realtek TX201 active. This is why I am investigating the platform/PCH crash log rather than treating the failure as specific to one NIC model.

PMC record

The PMC record is recognized as:

Record Type : PMC
Revision    : 5
Product ID  : 0x037
Size        : 2508 bytes
Checksum    : Valid
Reason      : 0x00030010

The decoded header also indicates that record collection completed successfully.

PMC_RST raw data

The 40-byte PMC_RST record is:

00000000  01 71 03 17 0A 00 00 00 02 18 00 00 02 00 00 00
00000010  00 00 00 00 02 00 00 00 00 00 00 00 EF BE AD DE
00000020  EF BE AD DE EF BE AD DE

I have deliberately not assigned register or bit meanings from another product's collateral because the mapping for Product ID 0x037 is not public and may differ.

Request

Could you please:

Confirm which platform/PCH Product ID 0x037 corresponds to.
Add or provide the matching collateral for at least:
decode-defs/PMC/5/layout.csv
decode-defs/PMC_TRACE/1/... if publishable
decode-defs/PMC_RST/1/layout.csv
the corresponding target_info.json
If PMC_TRACE decoding cannot be published, would you be able to decode the binary internally or forward it to the appropriate team, as was done for issue #90?
If possible, identify what PMC reason 0x30010 represents on this product.

The underlying failure is a complete system hard freeze with no useful final kernel journal entry immediately before the hang, so the PMC / PMC_TRACE payload may be the best evidence available.

I have the following binary blobs available and can attach them:

original saved Boot Error Region
PMC region: 2508 bytes
PMC_TRACE region: 512 bytes
PMC_RST region: 40 bytes

Thank you.

BERR-freeze-2026-08-22-crashlogs.zip

[scoommy]

Second independent freeze with the same PMC reason

I have now captured a second independent hard freeze and preserved a new ACPI BERT record.

This second BERT is not a copy of the first one:

• First BERT SHA256: 6839b33dd3adae620d9bd227dc34cf86d94d85d100799a2154153a439c7205aa
• Second BERT SHA256: 8e1f1023181eb54360898f07d8f818130ae6245f754afa17a8a5ffbee3e883c2

iclg info reports the same record structure again:

  #   Record Type  Rev.  Product  Size  Skt  Checksum  Die
----- ------------ ----- -------- ----- ---- --------- ----
 0-0  PMC              5 0x037     2508    0 Valid
 1-0  PMC_TRACE        1 0x037      512    0
 2-0  PMC_RST          1 0x037       40    0

And iclg triage again reports exactly:

CRASHLOG_REASON.PMC.30010H

I unpacked both BERT records and compared the individual regions:

PMC region: different between the two freezes
PMC_TRACE region: different between the two freezes
PMC_RST region: bit-for-bit identical between the two freezes

The identical PMC_RST result was verified by SHA256 comparison, while the main PMC and PMC_TRACE payloads have different hashes.

An additional potentially useful detail: during this second freeze, PCIe ASPM was disabled in the motherboard BIOS and Linux was also booted with:

pcie_aspm=off

The system nevertheless experienced the same complete hard freeze and produced the same PMC.30010H reason.

All four onboard Intel I226-V controllers were also disabled in BIOS during this test; the active network adapter was the add-in Realtek RTL8125.

This makes the repeated PMC.30010H signature particularly interesting, as it is reproducible across separate freezes despite the additional PCIe/ASPM isolation.

I am attaching the three unpacked records from this second freeze:

PMC rev 5, 2508 bytes
PMC_TRACE rev 1, 512 bytes
PMC_RST rev 1, 40 bytes

BERR-freeze-2026-08-22-121711-crashlogs.zip