Skip to contentSkip to conversation

"io mapaddr 0xXXXXX not valid" in smc-mca in 2.4.x

DISCUSSION Stephen Mollett 3 participants 4 messages Feb 26, 2001 – Feb 27, 2001

lore
Share & tools
lore
#0 lore raw

In kernel version 2.4.x (x from 0 to 2-ac3), the smc-mca driver gives many errors like the following on the console log:

io mapaddr 0xXXXXX not valid at smc-mca.c:YYY!

where XXXXX is an address within the shared-memory assigned to the adapter card, and YYY is 378, 398 or 408.

I have tested the driver on two IBM MCA systems - a 9577 and an 8590. I have tried three different network adapters in each machine:

WD Ethercard PLUS 10T/A (WD8003W/A)
SMC Ethercard PLUS Elite/A BNC/AUI (WD8013EP/A) SMC Ethercard PLUS Elite/A UTP/AUI (WD8013WP/A)

All the adapters give the errors, in both machines. All adapters and both PS/2s are known to be good with kernel 2.2.17.

* please cc any responses to me <[email protected]> as I do not subscribe to the list.

Regards,
Stephen Mollett

__________________________________________________
Do You Yahoo!?
Get email at your own domain with Yahoo! Mail.
http://personal.mail.yahoo.com/

#1 lore raw
Quoted 10 lines · On 02.26 Stephen Mollett wrote:

In kernel version 2.4.x (x from 0 to 2-ac3), the smc-mca driver gives many errors like the following on the console log:

io mapaddr 0xXXXXX not valid at smc-mca.c:YYY!

where XXXXX is an address within the shared-memory assigned to the adapter card, and YYY is 378, 398 or 408.

AFAIK, it is a bug in the driver. That output is generated by arch/i386/lib/iodebug.c, a routine to catch buggy drivers like that. Look at
include/asm-i386/io.h, line 112 or so:
/*
* Temporary debugging check to catch old code using
* unmapped ISA addresses. Will be removed in 2.4.
*/
#if CONFIG_DEBUG_IOVIRT
extern void *__io_virt_debug(unsigned long x, const char *file, int line); extern unsigned long __io_phys_debug(unsigned long x, const char *file, int li
ne);
#define __io_virt(x) __io_virt_debug((unsigned long)(x), __FILE__, __LINE__)
//#define __io_phys(x) __io_phys_debug((unsigned long)(x), __FILE__, __LINE__) #else
#define __io_virt(x) ((void *)(x))
//#define __io_phys(x) __pa(x)
#endif

So you should mail driver mantainer, if it is known....

Signature · 4 lines
J.A. Magallon                                                      $> cd pub
mailto:[email protected]                                          $> more beer

Linux werewolf 2.4.2-ac4 #2 SMP Mon Feb 26 00:21:23 CET 2001 i686
#2 lore raw
Quoted 4 lines

AFAIK, it is a bug in the driver. That output is generated by arch/i386/lib/iodebug.c, a routine to catch buggy drivers like that. Look at
include/asm-i386/io.h, line 112 or so:

Yes. Its someone using readb and friends on an unremapped address. In this case I had a look and a guess at a fix is in ac5 (out RSN)

#3 lore raw
Quoted 1 line · Alan Cox

... I had a look and a guess at a fix is in ac5 (out RSN)

You'll be glad to hear that the driver is now working perfectly. Thanks!

Regards,
Stephen Mollett