Re: PCI arbiter crash on last qemu image

[Damien Zammit]

On 18/8/20 6:51 am, Joan Lledó wrote:
> El 17/8/20 a les 1:51, Damien Zammit ha escrit:
>> Perhaps a better way to fix the mapping problem I encountered
>> is by removing the check for previous mappings when trying to map regions,
> 
> I could check the pointer before reading from it at func_files.c, and if
> it happens to be null, then call the libpciaccess mapping function from
> there, i.e. mapping the memory in the first access instead of doing it
> during the startup. So there's no need to make any changes in your
> patch, do you think that'd work?
> 

That would probably work, but I don't want to break other usages of pciaccess
by including my change upstream, doing it your way would mean every other use 
case
for pciaccess would need to do the same thing as you are suggesting.

Perhaps we can just remove the check instead after all in pciaccess,
as that would be compatible with existing code.

Damien