FWIW, I've seen strange but hard-to-reproduce behavior with the PCMCIA
stuff in test6-pre6 and test6-pre9 on my laptop (Sony VAIO 505TX). I'm
primarily using the wavelan_cs module, which may have something to do
with the problems (wavelan_cs seems to behave more poorly than the other
network drivers I've used on this laptop). Here are the things I've
seen:
* wavelan_cs sometimes doesn't get unloaded when the card is removed
* after some time, the card services seem to get wedged, and don't find
any cards when I insert or remove them. (restarting cardmgr doesn't
help)
* I got an OOPS while doing "insmod wavelan_cs" at one point. I forgot
to save the current ksyms, though, so I couldn't get anything
interesting from it.
I hadn't bothered to send a note to LKML up to this point, b/c I haven't
been able to consistently reproduce any of these problems. Given that
others are having problems as well, though, perhaps these additional
clues will be meaningful to someone. I'll send a followup note if I
find any additional information (or if the problems seem to go away with
test6, which I haven't tried yet....)
Best,
-Rob von Behren
John Kennedy wrote:
>
> On Thu, Aug 10, 2000 at 01:37:22AM -0700, David Ford wrote:
> > David Ford wrote:
> > Hmm, not so nifty. test6 completely breaks pcmcia with either package. The
> > kernel pcmcia modules load and all is configured but no packets are passed and
> > netdev timeouts and status errors as well as a supposed lack of eeprom. the
> > top socket still hangs the machine.
>
> I happened to be playing with 2.4.0-test6 on my laptop last night
> and your email caught my eye. For me, test6 wasn't totally broken.
> There does seem to be a mandatory 60-SECOND DELAY starting up any
> network connection though (even myself, external IP or loopback).
> Once the connection starts up (SSH, pinging some box, etc) it stays up
> and seems fine so far.
>
> In my case, I'm using the kernel pcmcia modules and a slightly patched
> version of xirc2ps_cs.o out of pcmcia-cs-3.1.17.
>
> [#define enet_statistics net_device_stats]
>
> This kernel was installed on a base system that is also running
> vmlinux-2.2.17pre15 where pcmcia-cs compiled cleanly and was fully
> installed, so I'm sure there is some amount of fudge going on here.
>
> > dhinds Aug.7 package compiles more cleanly than .19 but fails to load, attached
> > are snippets from this.
>
> I'll have to look into those.
>
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> Linux version 2.4.0-test6 (warlock@pandora) (gcc version 2.95.2 19991024 (release)) #135 SMP Wed Aug 9 23:39:18 PDT 2000
> ...
> Aug 10 07:58:32 src@pandora cardmgr[139]: starting, version is 3.1.17
> Aug 10 07:58:32 src@pandora cardmgr[139]: watching 2 sockets
> Aug 10 07:58:32 src@pandora cardmgr[139]: Card Services release does not match
> Aug 10 07:58:32 src@pandora kernel: cs: IO port probe 0x0c00-0x0cff: clean.
> Aug 10 07:58:32 src@pandora kernel: cs: IO port probe 0x0800-0x08ff: clean.
> Aug 10 07:58:32 src@pandora kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x220-0x22f 0x330-0x337 0x388-0x38f 0x398-0x39f 0x4d0-0x4d7
> Aug 10 07:58:32 src@pandora kernel: cs: IO port probe 0x0a00-0x0aff: clean.
> Aug 10 07:58:32 src@pandora cardmgr[139]: initializing socket 0
> Aug 10 07:58:32 src@pandora kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean.
> Aug 10 07:58:32 src@pandora cardmgr[139]: socket 0: Xircom CE3-10/100 Fast Ethernet
> Aug 10 07:58:32 src@pandora cardmgr[139]: executing: 'modprobe xirc2ps_cs'
> Aug 10 07:58:32 src@pandora cardmgr[139]: + modprobe: Can't locate module xirc2ps_cs
> Aug 10 07:58:32 src@pandora cardmgr[139]: modprobe exited with status 255
> Aug 10 07:58:32 src@pandora cardmgr[139]: executing: 'insmod /lib/modules/2.4.0-test6/pcmcia/xirc2ps_cs.o'
> Aug 10 07:58:32 src@pandora kernel: xirc2ps_cs.c 1.31 1998/12/09 19:32:55 (dd9jn+kvh)
> Aug 10 07:58:32 src@pandora cardmgr[139]: executing: './network start eth0'
> Aug 10 07:58:32 src@pandora kernel: eth0: Xircom: port 0x300, irq 3, hwaddr 00:10:A4:C3:59:02
> Aug 10 07:58:35 src@pandora kernel: eth0: MII link partner: 0021
> Aug 10 07:58:35 src@pandora kernel: eth0: MII selected
> Aug 10 07:58:35 src@pandora kernel: eth0: media 10BaseT, silicon revision 4
> ...
>
> -
> 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/
-
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 : Tue Aug 15 2000 - 21:00:21 EST