Created attachment 262530 [details] patch to recognize 0x91a3 PCI device ID for Marvell 9128 chip I have an old Gigabyte GA-P55A-UD4 motherboard. Two of the SATA slots are controlled by a Marvell 9128 chip. This chip is not recognized on boot by FreeBSD (14.3-p3), so those SATA slots are unusable. The reason it's not recognized is that it has an unexpected PCI ID. I included a small patch to have the kernel recognize this ID, and it is working locally. Similar code was added to Linux in 2010 (50be5e3657...), so I can't be the only one (: Note: I am not very familiar with the FreeBSD kernel. In particular, I copied the "quirks" value from the other existing Marvell 9128 entry. I am just guessing that is correct, but not 100% sure. For reference, this is the relevant bit from 'pciconv -lv': ahci0@pci0:2:0:0: class=0x01018f rev=0x11 hdr=0x00 vendor=0x1b4b device=0x91a3 subvendor=0x1458 subdevice=0xb000 vendor = 'Marvell Technology Group Ltd.' class = mass storage subclass = ATA
correction: I am on 14.2-p3, not .3 (though I doubt it matters, for this)
^Triage: now that Bugzilla supports the Patch flag for attachments, the [patch] convention is obsolete.
Created attachment 262755 [details] patch to recognize 0x91a3 PCI device ID for Marvell 9128 chip Re-added this patch, with the "patch" checkbox checked. Not sure from the last comment if that's required. The patch is unchanged. Also, it has been running successfully for me for a while, if that helps give it some weight.
I like to commit this, but need to do some research first. Can you provide a picture of the chip? The web gives me slightly other info and Copilot says: Marvell 88SE9123 SATA 6Gb/s Controlle dmesg output would good or something where vendor and device id are visible.
(In reply to Michael Osipov from comment #4) > dmesg output would good or something where vendor and device id are visible. The vendor/device ID should be the ones from the 'pciconv -lv' output in the original bug description, no? In particular (copied from above): ... vendor=0x1b4b device=0x91a3 subvendor=0x1458 subdevice=0xb000 But let me know if you need something else. > Can you provide a picture of the chip? I should be able to get a photograph — will add in a bit. The chip does have printed on it "88SE9128-NAA2" (see patch comments), but I will try to get a full picture, too.
Created attachment 262790 [details] Image of chip Photograph of chip. (sorry about blown-out colors; had to use flash to get it. Adjusted contrast to make more legible)
Also, in case it helps, here is product info for the mobo itself. It makes various mention of the Marvell chip: * main page: https://www.gigabyte.com/Motherboard/GA-P55A-UD4-rev-20 * PDF, user manual (English): https://download1.gigabyte.com/Files/Manual/mb_manual_ga-p55a-ud4(p)_v2.0_e.pdf?v=621fdca7651d5b9403fa8316361bb991
I just upgraded to 14.3, and the patch seems to be working there, too. I updated the version on this bug, but I guess it really applies to multiple versions; not sure the best way to express that in the bug metadata.
My bad, took me too long to respond. Here are two more references: * https://forum-en.msi.com/index.php?threads/big-bang-xpower-marvell-9128-firmware-issue.156443/ * https://www.hardwareluxx.de/community/threads/problem-marvell-88se9128-auf-asus-p6x58d-e-erkennt-nicht-dvd-laufwerk-sh-b083l.799977/ Obviously, Marvell chips appeared unter millions of devids depending on the board vendor.
https://reviews.freebsd.org/D56464
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=c72edc2b2eb283450df8db05cf9321aee7de0bcb commit c72edc2b2eb283450df8db05cf9321aee7de0bcb Author: John W <jwdevel@gmail.com> AuthorDate: 2025-07-28 05:17:32 +0000 Commit: Michael Osipov <michaelo@FreeBSD.org> CommitDate: 2026-04-23 16:41:19 +0000 ahci(4): Add another device ID for Marvell 9128 SATA controller: 0x91a3 This was encountered on a Gigabyte GA-P55A-UD4 motherboard identifying itself as: ahci0@pci0:2:0:0: class=0x01018f rev=0x11 hdr=0x00 vendor=0x1b4b device=0x91a3 subvendor=0x1458 subdevice=0xb000 vendor = 'Marvell Technology Group Ltd.' class = mass storage subclass = ATA The physical chip has "88SE9128-NAA2" printed on it. Similar code has been in Linux for a long time: https://github.com/torvalds/linux/commit/50be5e3657cd2851a297dc0b3fd459f25829d29b Co-authored-by: Michael Osipov <michaelo@FreeBSD.org> PR: 288526 Reviewed by: jhb MFC after: 1 week Differential Revision: https://reviews.freebsd.org/D56464 sys/dev/ahci/ahci_pci.c | 2 ++ 1 file changed, 2 insertions(+)
A commit in branch stable/15 references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=bcef4427a470274ba1f0c5fa7affe41253a93633 commit bcef4427a470274ba1f0c5fa7affe41253a93633 Author: John W <jwdevel@gmail.com> AuthorDate: 2025-07-28 05:17:32 +0000 Commit: Michael Osipov <michaelo@FreeBSD.org> CommitDate: 2026-04-30 07:51:30 +0000 ahci(4): Add another device ID for Marvell 9128 SATA controller: 0x91a3 This was encountered on a Gigabyte GA-P55A-UD4 motherboard identifying itself as: ahci0@pci0:2:0:0: class=0x01018f rev=0x11 hdr=0x00 vendor=0x1b4b device=0x91a3 subvendor=0x1458 subdevice=0xb000 vendor = 'Marvell Technology Group Ltd.' class = mass storage subclass = ATA The physical chip has "88SE9128-NAA2" printed on it. Similar code has been in Linux for a long time: https://github.com/torvalds/linux/commit/50be5e3657cd2851a297dc0b3fd459f25829d29b Co-authored-by: Michael Osipov <michaelo@FreeBSD.org> PR: 288526 Reviewed by: jhb MFC after: 1 week Differential Revision: https://reviews.freebsd.org/D56464 (cherry picked from commit c72edc2b2eb283450df8db05cf9321aee7de0bcb) sys/dev/ahci/ahci_pci.c | 2 ++ 1 file changed, 2 insertions(+)
A commit in branch stable/14 references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=487eeda542a774ba9237d8bda58c9fcb99b2afba commit 487eeda542a774ba9237d8bda58c9fcb99b2afba Author: John W <jwdevel@gmail.com> AuthorDate: 2025-07-28 05:17:32 +0000 Commit: Michael Osipov <michaelo@FreeBSD.org> CommitDate: 2026-04-30 07:52:30 +0000 ahci(4): Add another device ID for Marvell 9128 SATA controller: 0x91a3 This was encountered on a Gigabyte GA-P55A-UD4 motherboard identifying itself as: ahci0@pci0:2:0:0: class=0x01018f rev=0x11 hdr=0x00 vendor=0x1b4b device=0x91a3 subvendor=0x1458 subdevice=0xb000 vendor = 'Marvell Technology Group Ltd.' class = mass storage subclass = ATA The physical chip has "88SE9128-NAA2" printed on it. Similar code has been in Linux for a long time: https://github.com/torvalds/linux/commit/50be5e3657cd2851a297dc0b3fd459f25829d29b Co-authored-by: Michael Osipov <michaelo@FreeBSD.org> PR: 288526 Reviewed by: jhb MFC after: 1 week Differential Revision: https://reviews.freebsd.org/D56464 (cherry picked from commit c72edc2b2eb283450df8db05cf9321aee7de0bcb) sys/dev/ahci/ahci_pci.c | 2 ++ 1 file changed, 2 insertions(+)
A commit in branch stable/13 references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=65d75038f32ca6a0e3e07aadb25bb4c89e7998ff commit 65d75038f32ca6a0e3e07aadb25bb4c89e7998ff Author: John W <jwdevel@gmail.com> AuthorDate: 2025-07-28 05:17:32 +0000 Commit: Michael Osipov <michaelo@FreeBSD.org> CommitDate: 2026-04-30 07:53:10 +0000 ahci(4): Add another device ID for Marvell 9128 SATA controller: 0x91a3 This was encountered on a Gigabyte GA-P55A-UD4 motherboard identifying itself as: ahci0@pci0:2:0:0: class=0x01018f rev=0x11 hdr=0x00 vendor=0x1b4b device=0x91a3 subvendor=0x1458 subdevice=0xb000 vendor = 'Marvell Technology Group Ltd.' class = mass storage subclass = ATA The physical chip has "88SE9128-NAA2" printed on it. Similar code has been in Linux for a long time: https://github.com/torvalds/linux/commit/50be5e3657cd2851a297dc0b3fd459f25829d29b Co-authored-by: Michael Osipov <michaelo@FreeBSD.org> PR: 288526 Reviewed by: jhb MFC after: 1 week Differential Revision: https://reviews.freebsd.org/D56464 (cherry picked from commit c72edc2b2eb283450df8db05cf9321aee7de0bcb) sys/dev/ahci/ahci_pci.c | 2 ++ 1 file changed, 2 insertions(+)