Re: [Qemu-devel] [PATCH 2/4] cputlb: move TLB handling to a separate file

[Andreas Färber]

Am 28.04.2012 13:48, schrieb Blue Swirl:
> On Sun, Apr 22, 2012 at 15:35, Blue Swirl <address@hidden> wrote:
>> diff --git a/cputlb.c b/cputlb.c
>> new file mode 100644
>> index 0000000..b7d8f07
>> --- /dev/null
>> +++ b/cputlb.c
[...]
>> +void tlb_reset_dirty_range(CPUTLBEntry *tlb_entry, uintptr_t start,
>> +                           uintptr_t length)
> 
> This...
> 
>> +{
>> +    uintptr_t addr;
>> +
>> +    if (tlb_is_dirty_ram(tlb_entry)) {
>> +        addr = (tlb_entry->addr_write & TARGET_PAGE_MASK) + 
>> tlb_entry->addend;
>> +        if ((addr - start) < length) {
>> +            tlb_entry->addr_write |= TLB_NOTDIRTY;
>> +        }
>> +    }
>> +}
[...]
>> diff --git a/cputlb.h b/cputlb.h
>> new file mode 100644
>> index 0000000..d16c22e
>> --- /dev/null
>> +++ b/cputlb.h
[...]
>> +void tlb_reset_dirty_range(CPUTLBEntry *tlb_entry, unsigned long start,
>> +                           unsigned long length);
> 
> ... doesn't match this prototype, so build would fail on win32.
> 
> Maybe this is 1.2 material anyway.

Personally I'd like to have those big code movements sorted out earlier
than later (including any outstanding AREG0 refactorings) due to my
CPUArchState -> CPUState consolidation work.

So if you and Richard are happy with the code changes except for this
mismatch then I can check both mingw cross builds if you name a branch.

But I'd also be fine with a semiofficial "next" branch for 1.2.

Andreas

-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg