Further to my problem with failure at probe time of my PCI NE2000 clone (and silly previous proposal re call to eth_pio_write at line 880 of ns8390.c), here is a possible fix - hopefully not so silly. Test whether probing a PCI board and, if so, go immediately into 16 bit mode before the first memory test, i.e. *** ns8390.c.orig Fri Mar 17 07:53:06 2000 --- ns8390.c Fri Mar 17 07:54:14 2000 *************** *** 877,882 **** --- 877,885 ---- outb(D8390_DCR_FT1 | D8390_DCR_LS, eth_nic_base + D8390_P0_DCR); outb(MEM_8192, eth_nic_base + D8390_P0_PSTART); outb(MEM_16384, eth_nic_base + D8390_P0_PSTOP); + #ifdef INCLUDE_NEPCI + eth_flags |= FLAG_16BIT; + #endif eth_pio_write(test, 8192, sizeof(test)); eth_pio_read(8192, testbuf, sizeof(test)); if (!memcmp(test, testbuf, sizeof(test))) It seems that in the kernel driver (ne2k-pci.c) it is assumed that all PCI boards operate (at least) in 16 bit mode so the above patch just makes the same assumption. I don't know (and can't test) if this may raise problems if a machine has both PCI and ISA boards installed and an ISA board is 8390 based. chris _______________________________________________________________ Chris Willing Telephone (61-2) 9351 3005 VisLab, A28 Facsimile (61-2) 9351 1880 University of Sydney mailto:firstname.lastname@example.org NSW 2006 Australia http://www.vislab.usyd.edu.au =========================================================================== This Mail was sent to netboot mailing list by: Christoph Willing <email@example.com> To get help about this list, send a mail with 'help' as the only string in it's body to firstname.lastname@example.org. If you have problems with this list, send a mail to email@example.com.
For requests or suggestions regarding this mailing list archive please write to firstname.lastname@example.org.