On 01/15/2018 09:31 AM, Richard Henderson wrote:
> But we can't *not* swap the arguments to the generated comparison. Otherwise
> we're computing the wrong thing.
Bah. I have of course forgotten about the RSC instruction, which takes an
immediate. Will fix.
r~
On 01/15/2018 06:27 AM, Peter Maydell wrote:
> We could avoid the annoying "load LE/GE immediates to tempreg"
> extra code by having tcg_out_cmp2() return a flag to tell
> the caller which way round to put the conditions for its two
> conditional ARITH_MOV insns (for setcond2) or which condition
>
On 01/15/2018 06:27 AM, Peter Maydell wrote:
> (may be worth mentioning
> in particular how "rIM" differs from "rINK" -- I think the answer is that
> we want to say "valid-immediate OR (valid-if-inverted AND valid-if-negated)",
> whereas 'rINK' is "valid-immediate OR valid-if-inverted OR valid-if-n
On 15 January 2018 at 16:18, Richard Henderson wrote:
> On 01/15/2018 06:27 AM, Peter Maydell wrote:
>> The code looks OK to me. I'm guessing the sparc guest failure is
>> just because we now generate more code for some of the comparison
>> cases and that doesn't fit in the buffer...
>
> I don't r
On 01/15/2018 06:27 AM, Peter Maydell wrote:
> On 10 January 2018 at 05:39, Richard Henderson wrote:
>> The code sequence we were generating was only good for unsigned
>> comparisons. For signed comparisions, use the sequence from gcc.
>>
>> Fixes booting of ppc64 firmware, with a patch changing
On 10 January 2018 at 05:39, Richard Henderson wrote:
> The code sequence we were generating was only good for unsigned
> comparisons. For signed comparisions, use the sequence from gcc.
>
> Fixes booting of ppc64 firmware, with a patch changing the code
> sequence for ppc comparisons.
>
> Signed
Quoting Richard Henderson (2018-01-09 23:39:08)
> The code sequence we were generating was only good for unsigned
> comparisons. For signed comparisions, use the sequence from gcc.
>
> Fixes booting of ppc64 firmware, with a patch changing the code
> sequence for ppc comparisons.
>
> Signed-off-
The code sequence we were generating was only good for unsigned
comparisons. For signed comparisions, use the sequence from gcc.
Fixes booting of ppc64 firmware, with a patch changing the code
sequence for ppc comparisons.
Signed-off-by: Richard Henderson
---
tcg/arm/tcg-target.inc.c | 112 +++