Re: [Qemu-devel] [PATCH 5/7] tcg-i386: Implement deposit operation.

[Alexander Graf <agraf@suse.de>]

Richard Henderson wrote:
> On 01/26/2011 08:50 AM, Alexander Graf wrote:
>   
>> Oh, you mean basically to have the following:
>>
>> TCGv_i32 regs32[16];
>> TCGv_i64 regs[16];
>>
>> Then declare both as globals with offset and just switch between the
>> access type using a disas struct variable. Once the TB ends, I'd
>> obviously have to sync back to 64 bit again.
>>     
>
> Maybe?  I don't think that regs32 can overlap with regs in the real
> cpu structure.  Otherwise you have the same sort of sync problems as
> was originally rejected.
>
> I had been picturing regs32 as a member of DisasContext, and it would
> hold tcg_temp_new_i32 variables as-needed.  Something like
>   

That would mean that during a TB the reg32 parts would be in
temporaries, right? So they'd end up being somewhere in a register.

What do I do on a page fault now? I could rerun the translator to find
out which registers would be stuck in a temporary, but I have no way to
actually read the temporary's value, as all register state is thrown
away on a page fault IIUC.

So the only alternative to this would be that I'd keep both 32 bit and
64 bit register states around in env, declare both as global, and sync
them manually on access. I guess.


Alex