On Mon, Jun 05, 2000 at 03:45:00AM -0700, Andre Hedrick wrote:
> On Mon, 5 Jun 2000, John D. Rowell wrote:
>
> > This is on a Toshiba Portege 3110CT laptop.
> >
> > tiny:~# lspci | grep IDE
> > 00:07.1 IDE interface: Intel Corporation 82440MX EIDE Controller
> >
> > Works fine with 2.2.14, 2.2.16pre7 and 2.3.99-pre[1-6], very stable
> > even with hdparm -u1 -c1 and idebus=66. Transfer rate by hdparm -tT
> > is about 11Mb/s.
>
> Your "idebus=66" is not *, but "idebus=33", and APM/ACPI may not play nice
> with unmasking the interrupt.
Yes, I was just trying to show how stable the old one was ;) I tried
it clean with no idebus and no hdparm tweaking, and get the same
results. The only thing that prevents the crashes is disabling DMA
(hdparm -d0), but DMA works fine with the other kernels _and_ is
detected and accepted by the kernel during boot as usual.
A quick peek at /usr/src/kernel-2.4.x/drivers/ide/piix.c tells me
that there are no settings for my specific chipset, but even
compiling with only the default kernel DMA support (no PIIX) gives
the same problems. My guess is that it is a timing related problem,
but I have no idea as of how to change the timings so I can try
them on my system.
Again, the same kernel configuration works fine up to 2.3.99-pre6.
I'm currently chugging along with 2.4.0-test1-ac8 with no DMA (about
5Mb/s transfer rate).
--jdrowell
-
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/
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:21 EST