Skip to contentSkip to conversation

Marvell SATA AHCI 0x6141

PATCH Davor 2 participants 4 messages Mar 9, 2007 – Mar 13, 2007 libata

lore
Share & tools
lore

Davor

author
#0 lore raw

HI

I have onboard
02:00.0 SATA controller: Marvell Technology Group Ltd. Unknown device 6141 (rev 01) 02:00.0 0106: 11ab:6141 (rev 01)

and it's an AHCI compatible device so a simple patch makes it work (patch agains linux-2.6.20.2)

ata/ahci.c renamed from ata/ahci.c.orig +3 -0
@@ -437,6 +437,9 @@ static const struct pci_device_id ahci_p
 	{ PCI_VDEVICE(SI, 0x1185), board_ahci }, /* SiS 966 */
 	{ PCI_VDEVICE(SI, 0x0186), board_ahci }, /* SiS 968 */
 
+	/* Marvell */
+	{ PCI_VDEVICE(MARVELL, 0x6141), board_ahci },	/* 6141 */
+
 	/* Generic, PCI class code for AHCI */
 	{ PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID,
 	  0x010601, 0xffffff, board_ahci },
#1 lore raw
Quoted 10 lines · Davor

HI

I have onboard 02:00.0 SATA controller: Marvell Technology Group Ltd. Unknown device 6141 (rev 01) 02:00.0 0106: 11ab:6141 (rev 01)

and it's an AHCI compatible device so a simple patch makes it work (patch agains linux-2.6.20.2)

ata/ahci.c static const struct pci_device_id ahci_p · line 441

Quoted 5 lines
ata/ahci.c renamed from ata/ahci.c.orig +2 -0
@@ -437,6 +437,9 @@ static const struct pci_device_id ahci_p
 	{ PCI_VDEVICE(SI, 0x1185), board_ahci }, /* SiS 966 */
 	{ PCI_VDEVICE(SI, 0x0186), board_ahci }, /* SiS 968 */
 
+	/* Marvell */
+	{ PCI_VDEVICE(MARVELL, 0x6141), board_ahci },	/* 6141 */

Yep, though you want the more-complete patch from libata-dev.git#mv-ahci or akpm's -mm.

Jeff

#2 lore raw

HI

For i386 it works (2.6.20.2).
(hotswap has bug,
disk has to first be inserted (nothing happens), then removed and insterted again
then it's recognized)

However for amd64 (2.6.20.2) on the same hardware it doesn't work. During boot the ata* messages appear on the ata4 controller where the disk is connected but the kernel waits long with some messages but in the end no disk is recognized. hoswap insert/remove doesn't help either to detect the disk on amd64

I'd like to try this patch if it would fix amd64 but How to download out the libata#mv-ahci patch?

Best regards, Davor

Quoted 29 lines · Jeff Garzik
Quoted 4 lines · Davor

HI

I have onboard
02:00.0 SATA controller: Marvell Technology Group Ltd. Unknown device 6141

(rev 01)

ata/ahci.c static const struct pci_device_id ahci_p · line 441

Quoted 13 lines

02:00.0 0106: 11ab:6141 (rev 01)

and it's an AHCI compatible device so a simple patch makes it work (patch agains linux-2.6.20.2)

ata/ahci.c renamed from ata/ahci.c.orig +2 -0
@@ -437,6 +437,9 @@ static const struct pci_device_id ahci_p
 	{ PCI_VDEVICE(SI, 0x1185), board_ahci }, /* SiS 966 */
 	{ PCI_VDEVICE(SI, 0x0186), board_ahci }, /* SiS 968 */
 
+	/* Marvell */
+	{ PCI_VDEVICE(MARVELL, 0x6141), board_ahci },	/* 6141 */

Yep, though you want the more-complete patch from libata-dev.git#mv-ahci or akpm's -mm.

Jeff

#3 lore raw

hi

I got it running on amd64 (with my minimalistic patch) I checked out the 2.6.21-rc3-mm there's some marvell related stuff but I think it's for a mixed marvell cards with serail and parallel ata. Still this patch has chip id 0x6145 while my 0x6141 (it's serial ata only) has to be added to get recognized by ahci.c

Quoted 52 lines · Davor Emard

HI

For i386 it works (2.6.20.2).
(hotswap has bug,
disk has to first be inserted (nothing happens), then removed and insterted again
then it's recognized)

However for amd64 (2.6.20.2) on the same hardware it doesn't work. During boot the ata* messages appear on the ata4 controller where the disk is connected but the kernel waits long with some messages but in the end no disk is recognized. hoswap insert/remove doesn't help either to detect the disk on amd64

I'd like to try this patch if it would fix amd64 but How to download out the libata#mv-ahci patch?

Best regards, Davor

Quoted 5 lines · Jeff Garzik
Quoted 4 lines · Davor

HI

I have onboard
02:00.0 SATA controller: Marvell Technology Group Ltd. Unknown device

6141

Quoted 24 lines

(rev 01)

ata/ahci.c static const struct pci_device_id ahci_p · line 441

Quoted 13 lines

02:00.0 0106: 11ab:6141 (rev 01)

and it's an AHCI compatible device so a simple patch makes it work (patch agains linux-2.6.20.2)

ata/ahci.c renamed from ata/ahci.c.orig +2 -0
@@ -437,6 +437,9 @@ static const struct pci_device_id ahci_p
 	{ PCI_VDEVICE(SI, 0x1185), board_ahci }, /* SiS 966 */
 	{ PCI_VDEVICE(SI, 0x0186), board_ahci }, /* SiS 968 */
 
+	/* Marvell */
+	{ PCI_VDEVICE(MARVELL, 0x6141), board_ahci },	/* 6141 */

Yep, though you want the more-complete patch from libata-dev.git#mv-ahci or akpm's -mm.

Jeff