Re: [Patch/ccmp] Cost instruction sequences to choose better expand order

2015-09-23 Thread Bernd Schmidt
No. Please see NOTE part of the description. AArch64 doesn't cost ccmp currently. It will be fixed by a seperate patch later. The testcase is thus marked as XFAIL. I'd prefer to do things in the right order. Your patch is approved, but please commit only after you can remove the xfail from the

Re: [Patch/ccmp] Cost instruction sequences to choose better expand order

2015-09-21 Thread Bernd Schmidt
On 09/18/2015 05:21 PM, Jiong Wang wrote: Current conditional compare (CCMP) support in GCC aim to optimize short circuit for cascade comparision, given a simple conditional compare candidate: if (a == 17 || a == 32) [...] The problem is current implementation always expand t0 first, then

Re: [Patch/ccmp] Cost instruction sequences to choose better expand order

2015-09-21 Thread Jiong Wang
Bernd Schmidt writes: > On 09/18/2015 05:21 PM, Jiong Wang wrote: >> >> Current conditional compare (CCMP) support in GCC aim to optimize >> short circuit for cascade comparision, given a simple conditional >> compare candidate: >> >>if (a == 17 || a == 32) > [...] >> The problem is current

Re: [Patch/ccmp] Cost instruction sequences to choose better expand order

2015-09-21 Thread pinskia
> On Sep 21, 2015, at 4:39 AM, Bernd Schmidt wrote: > >> On 09/18/2015 05:21 PM, Jiong Wang wrote: >> >> Current conditional compare (CCMP) support in GCC aim to optimize >> short circuit for cascade comparision, given a simple conditional >> compare candidate: >> >>

[Patch/ccmp] Cost instruction sequences to choose better expand order

2015-09-18 Thread Jiong Wang
Current conditional compare (CCMP) support in GCC aim to optimize short circuit for cascade comparision, given a simple conditional compare candidate: if (a == 17 || a == 32) it's represented like the following in IR: t0 = a == 17 t1 = a == 32 t2 = t0 || t1 Normally, CCMP contains