On Mon, 14 Jun 1999, Paul Jakma wrote:
> Hi,
>
> I'm trying to get 2.2.7-ac2 running on a Samsung UX board with onboard
> Symbios 53c875 scsi. But every time i boot it i get:
>
> <detects it as 53c875, irq,port, etc all correct>
> CACHE TEST FAILED: script execution failed.
> <and then:>
> CACHE INCORRECTLY CONFIGURED.
> <then it gives up, and the kernel can't mount it's root fs.
>
> This happens with both the sym53c8xx and ncr53c8xx drivers on 2.2.7-ac1 and
> ac2, but works fine with 2.2.9. (milo never has problems either). I've tried
> loads of different config options. Even copying the sym53c8x* files from
> linux-2.2.9/drivers/scsi over to 2.2.7-ac2 didn't work - exact same problem
> on boot.
So, it seems it is probably not a driver problem.
> I also downloaded the latest sym53c875 1.5 patches from ftp.tux.org but it
> fails to link vmlinux with undefined references to remap_pci, unmap_pci (and
> 1/2 other functions which i can't remember) in sym53c8xx.o. These functions
> are all contained within sym53c8xx.c. So why can't the linker find these
> functions in the object file? (yes i tried make clean and also mrproper).
>
> So what's up with 2.2.7-ac2 and sym53c875?? I need 2.2.7-ac for linux-raid.
> (2.2.9 is broken).
I will suggest you to try sym driver 1.5c that also will fail to compile
on Alpha, as it is. But with the following define added in sym53c8xx.h, it
has been reported to work fine on an Alpha platform.
#defined SCSI_NCR_PCI_MEM_NOT_SUPPORTED
I donnot remember which Linux version was successful, but your report let
me think that 2.2.7-ac2 will probably also fail with this driver version.
> Is there a solution to get the 1.3 to work with 2.2.7-ac? Or else how can i
> get the 1.5 version to link in, so i can test it? The hardware is fine, so
> it must be the kernel or the driver.
I dont know. And as long as I will not have an Alpha box to check the sym
and ncr drivers for Alpha, I will not be able to provide verified and
tested code for Alpha. Neither, I will be able to fix problems in the
kernel that may affect the drivers. Checking times the source code is
fine, but may fail sometimes.
Regards,
Gérard.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/