Hi Alex,
thanks for clarifying.
There's a number of reasons why this might fail:
- card really needs the reset before even accepting the probe I/O,
- PCMCIA interface needs to be enabled before the probe (pcmcia_init()
normally does that, but the card reset happens right at the start there),
- a probe read on a 16 bit card in 8 bit mode does appear to give the
correct result.
I'll have to come up with something better - the CISTPL_CFTABLE_ENTRY
config tuple ought to contain information about the IO path width
somewhere. Just a matter of parsing it out without duplicating all of
drivers/pcmcia/cistpl.c ... Any takers?
Cheers,
Michael
On 21/06/21 9:55 am, ALeX Kazik wrote:
Hi Michael,
thanks for testing again
I'm happy to help.
so autoprobe (no boot option) did_not_ work?
Yes, it did not work - sorry.
If anyone is curious: I'm testing with linux 5.12.9 (was the latest
stable when the v1 appeared) and an about 5 year old installation on
my A1200 with 68030+FPU@50Mhz and 64MB ram.
Alex.
