> On December 23, 2019 at 8:12 AM Mike Larkin <[email protected]> wrote:
>
>
> Can you configure the SATA interface on this (very old) machine to use AHCI
> instead of IDE?
>
> -ml
>
Thank you for your answer.
Sadly this BIOS does not allow the possibility to choose AHCI.
In the @misc mailing list I have found a way that should enforce the
kernel to use AHCI regardless of the BIOS settings adding these lines to
ahci_pci.c:
@@ -80,6 +80,8 @@
struct pci_attach_args *);
static const struct ahci_device ahci_devices[] = {
+ { PCI_VENDOR_NVIDIA, PCI_PRODUCT_NVIDIA_MCP51_SATA,
+ NULL, NULL },
With no good results:
*The SATA SSD Micron 5200 continue to have the same problem
*The other working SATA HD is disappeared
*New message in dmesg:
ahci0 at pci0 dev 14 function 0 "NVIDIA MCP51 SATA" rev 0xa1:
apic 4 int 20, unsupported AHCI revision 0x00000000
I have checked in the dmesg of Linux, FreeBSD, DragonFlyBSD and there is
not any message related to AHCI, it seems possible that they work in
IDE mode with this SSD.
At this point I start to believe that the MCP51 SATA controller does not
support AHCI.
Would be nice to see this SSD works in IDE mode also with OpenBSD but it
is not so important until I can circumvent the problem, so that the
developers can focus on more important areas of the OS.
Do you think that with the use of an expansion card can I bypass the
problem?
I read that the Marvell 88SE9128 controller works fine with OpenBSD and
in AHCI mode.
