Re: IDE disks in Amigas
On Fri, 4 Apr 2008, Ingo Juergensmann wrote:
> Geert Uytterhoeven wrote:
> >> Under AmigaOS the performance of the IDE disk was "OK", but under Linux
> >> it's
> >> quite slow. hdparm shows all options OFF and trying to set DMA to on
> >> resulted in an error. Is there a reason to not being able to set DMA on
> >> IDE
> >> for Amigas or is it a bug in hdparm or something else?
> > The Amiga IDE hardware cannot do DMA, only PIO.
>
> *yawn*
>
> > Setting the multiple sector count (`hdparm -m') may help, though.
>
> Will try...
>
> > What figures do you get from `hdparm -tT /dev/hda'?
>
> elgar:~# hdparm -tT /dev/hda
>
> /dev/hda:
> Timing cached reads: 26 MB in 2.01 seconds = 12.93 MB/sec
> Timing buffered disk reads:
> Message from syslogd@elgar at Fri Apr 4 12:44:15 2008 ...
> elgar kernel: Bad page state in process 'hdparm'
>
> Message from syslogd@elgar at Fri Apr 4 12:44:15 2008 ...
> elgar kernel: page:002f0f34 flags:0x00000004 mapping:00000000
> mapcount:-2147483648 count:0
>
> Message from syslogd@elgar at Fri Apr 4 12:44:15 2008 ...
> elgar kernel: Trying to fix it up, but a reboot is needed
>
> Message from syslogd@elgar at Fri Apr 4 12:44:15 2008 ...
> elgar kernel: Backtrace:
> 4 MB in 4.06 seconds = 1008.11 kB/sec
Hmm... Indeed not so fast. As you have an '060 and I have an '040 on the
notorious A3640 CPU card, I'd expect higher numbers (at least 2 MiB/s).
Does it improve without NCR_700 noise?
> Apr 4 12:44:15 elgar kernel: Bad page state in process 'hdparm'
> Apr 4 12:44:15 elgar kernel: page:002f0f34 flags:0x00000004
> mapping:00000000 ma
> pcount:-2147483648 count:0
> Apr 4 12:44:15 elgar kernel: Trying to fix it up, but a reboot is needed
> Apr 4 12:44:15 elgar kernel: Backtrace:
> Apr 4 12:44:15 elgar kernel: Call Trace: [<0004766e>] bad_page+0x52/0x76
> Apr 4 12:44:15 elgar kernel: [<0004791c>]
> get_page_from_freelist+0x28a/0x328
> Apr 4 12:44:15 elgar kernel: [<000201d0>] _060_fpsp_effadd+0xca8c/0xd518
> Apr 4 12:44:15 elgar kernel: [<0004827c>] __alloc_pages+0x4a/0x262
> Apr 4 12:44:15 elgar kernel: [<000201d0>] _060_fpsp_effadd+0xca8c/0xd518
> Apr 4 12:44:15 elgar last message repeated 2 times
> Apr 4 12:44:15 elgar kernel: [<00049c62>]
> __do_page_cache_readahead+0x128/0x2a
> 6
> Apr 4 12:44:16 elgar kernel: [<000201d0>] _060_fpsp_effadd+0xca8c/0xd518
> Apr 4 12:44:16 elgar kernel: [<000438de>] find_get_page+0x0/0x50
> Apr 4 12:44:16 elgar kernel: [<00003ebe>] buserr_c+0x4e2/0x6a4
> Apr 4 12:44:16 elgar kernel: [<00001000>] _stext+0x0/0x1000
> Apr 4 12:44:16 elgar kernel: [<00160000>] NCR_700_queuecommand+0x400/0x47e
^^^^^^^^^^^^^^^^^^^^
That's not the Amiga `Gayle' IDE driver ;-)
> That's with kernel 2.6.23-m68k which is running w/o problems on my other
> Amigas. I get this kind of errors when I'm attaching a SCSI disk to the
> SCSI module of the CyberStorm1:
Sorry, I'm afraid I can't help you with that.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Reply to: