Displaying 16 results from an estimated 16 matches for "sqrtpd".
2013 Jul 19
0
[LLVMdev] llvm.x86.sse2.sqrt.pd not using sqrtpd, calling a function that modifies ECX
...e that I've been getting, both with
> CodeGenOpt::Default and CodeGenOpt::None . The crash isn't occurring with
> CodeGenOpt::None, but that seems to be because ECX isn't being used - it
> still gets set to 0x7fffffff by one of the calls to 76719BA1
>
> I notice that X86::SQRTPD[m|r] appear in X86InstrInfo::isHighLatencyDef. I
> was thinking an optimization might be removing it, but I don't get the
> sqrtpd instruction even if the createJIT optimization level turned off.
>
> I am trying this with the Release 3.3 code - I'll try it with trunk and
> se...
2013 Jul 19
3
[LLVMdev] fptoui calling a function that modifies ECX
...[(X86WinFTOL RFP64:$src)]>,
Requires<[In32BitMode]>;
}
On Thu, Jul 18, 2013 at 11:59 PM, Peter Newman <peter at uformia.com> wrote:
> Oh, excellent point, I agree. My bad. Now that I'm not assuming those
> are the sqrt, I see the sqrtpd's in the output. Also there are three
> fptoui's and there are 3 call instances.
>
> (Changing subject line again.)
>
> Now it looks like it's bug #13862
>
> On 19/07/2013 4:51 PM, Craig Topper wrote:
>
> I think those calls correspond to this
>
> %11...
2013 Jul 19
2
[LLVMdev] fptoui calling a function that modifies ECX
...FP64:$src)]>,
> Requires<[In32BitMode]>;
> }
>
>
> On Thu, Jul 18, 2013 at 11:59 PM, Peter Newman <peter at uformia.com> wrote:
>
>> Oh, excellent point, I agree. My bad. Now that I'm not assuming those
>> are the sqrt, I see the sqrtpd's in the output. Also there are three
>> fptoui's and there are 3 call instances.
>>
>> (Changing subject line again.)
>>
>> Now it looks like it's bug #13862
>>
>> On 19/07/2013 4:51 PM, Craig Topper wrote:
>>
>> I think those calls...
2013 Jul 19
0
[LLVMdev] fptoui calling a function that modifies ECX
Oh, excellent point, I agree. My bad. Now that I'm not assuming those
are the sqrt, I see the sqrtpd's in the output. Also there are three
fptoui's and there are 3 call instances.
(Changing subject line again.)
Now it looks like it's bug #13862
On 19/07/2013 4:51 PM, Craig Topper wrote:
> I think those calls correspond to this
>
> %110 = fptoui double %109 to i32
>
&g...
2013 Jul 19
2
[LLVMdev] fptoui calling a function that modifies ECX
...Requires<[In32BitMode]>;
>> }
>>
>>
>> On Thu, Jul 18, 2013 at 11:59 PM, Peter Newman <peter at uformia.com> wrote:
>>
>>> Oh, excellent point, I agree. My bad. Now that I'm not assuming those
>>> are the sqrt, I see the sqrtpd's in the output. Also there are three
>>> fptoui's and there are 3 call instances.
>>>
>>> (Changing subject line again.)
>>>
>>> Now it looks like it's bug #13862
>>>
>>> On 19/07/2013 4:51 PM, Craig Topper wrote:
>>&...
2013 Jul 19
0
[LLVMdev] fptoui calling a function that modifies ECX
...Requires<[In32BitMode]>;
> }
>
>
> On Thu, Jul 18, 2013 at 11:59 PM, Peter Newman <peter at uformia.com
> <mailto:peter at uformia.com>> wrote:
>
> Oh, excellent point, I agree. My bad. Now that I'm not assuming
> those are the sqrt, I see the sqrtpd's in the output. Also there
> are three fptoui's and there are 3 call instances.
>
> (Changing subject line again.)
>
> Now it looks like it's bug #13862
>
> On 19/07/2013 4:51 PM, Craig Topper wrote:
>> I think those calls correspond to th...
2013 Jul 19
0
[LLVMdev] fptoui calling a function that modifies ECX
...>
>>
>> On Thu, Jul 18, 2013 at 11:59 PM, Peter Newman <peter at uformia.com
>> <mailto:peter at uformia.com>> wrote:
>>
>> Oh, excellent point, I agree. My bad. Now that I'm not
>> assuming those are the sqrt, I see the sqrtpd's in the
>> output. Also there are three fptoui's and there are 3 call
>> instances.
>>
>> (Changing subject line again.)
>>
>> Now it looks like it's bug #13862
>>
>> On 19/07/2013 4:51 PM, Craig To...
2013 Jul 20
0
[LLVMdev] fptoui calling a function that modifies ECX
...n Thu, Jul 18, 2013 at 11:59 PM, Peter Newman
>>> <peter at uformia.com <mailto:peter at uformia.com>> wrote:
>>>
>>> Oh, excellent point, I agree. My bad. Now that I'm not
>>> assuming those are the sqrt, I see the sqrtpd's in the
>>> output. Also there are three fptoui's and there are 3
>>> call instances.
>>>
>>> (Changing subject line again.)
>>>
>>> Now it looks like it's bug #13862
>>>
>...
2013 Jul 19
4
[LLVMdev] SIMD instructions and memory alignment on X86
Hmm, I'm not able to get those .ll files to compile if I disable SSE and I
end up with SSE instructions(including sqrtpd) if I don't disable it.
On Thu, Jul 18, 2013 at 10:53 PM, Peter Newman <peter at uformia.com> wrote:
> Is there something specifically required to enable SSE? If it's not
> detected as available (based from the target triple?) then I don't think we
> enable it specifi...
2013 Jul 19
0
[LLVMdev] llvm.x86.sse2.sqrt.pd not using sqrtpd, calling a function that modifies ECX
...hing the compiled code that I've been getting, both with
CodeGenOpt::Default and CodeGenOpt::None . The crash isn't occurring
with CodeGenOpt::None, but that seems to be because ECX isn't being used
- it still gets set to 0x7fffffff by one of the calls to 76719BA1
I notice that X86::SQRTPD[m|r] appear in X86InstrInfo::isHighLatencyDef.
I was thinking an optimization might be removing it, but I don't get the
sqrtpd instruction even if the createJIT optimization level turned off.
I am trying this with the Release 3.3 code - I'll try it with trunk and
see if I get a differen...
2013 Jul 19
2
[LLVMdev] SIMD instructions and memory alignment on X86
...; 76719BD8 add ecx,7FFFFFFFh
> 76719BDE sbb eax,0
> 76719BE1 mov edx,dword ptr [esp+14h]
> 76719BE5 sbb edx,0
> 76719BE8 leave
> 76719BE9 ret
>
>
> As you can see at 76719BD5, it modifies ECX .
>
> I don't know that this is the sqrtpd function (for example, I'm not seeing
> any SSE instructions here?) but whatever it is, it's being called from the
> IR I attached earlier, and is modifying ECX under some circumstances.
>
>
> On 19/07/2013 3:29 PM, Craig Topper wrote:
>
> That should map directly to sq...
2013 Jul 19
0
[LLVMdev] SIMD instructions and memory alignment on X86
...7FFFFFFFh
> 76719BDE sbb eax,0
> 76719BE1 mov edx,dword ptr [esp+14h]
> 76719BE5 sbb edx,0
> 76719BE8 leave
> 76719BE9 ret
>
>
> As you can see at 76719BD5, it modifies ECX .
>
> I don't know that this is the sqrtpd function (for example, I'm
> not seeing any SSE instructions here?) but whatever it is, it's
> being called from the IR I attached earlier, and is modifying ECX
> under some circumstances.
>
>
> On 19/07/2013 3:29 PM, Craig Topper wrote:
>> That s...
2013 Jul 19
0
[LLVMdev] SIMD instructions and memory alignment on X86
...[esp]
76719BD5 mov ecx,dword ptr [esp]
76719BD8 add ecx,7FFFFFFFh
76719BDE sbb eax,0
76719BE1 mov edx,dword ptr [esp+14h]
76719BE5 sbb edx,0
76719BE8 leave
76719BE9 ret
As you can see at 76719BD5, it modifies ECX .
I don't know that this is the sqrtpd function (for example, I'm not
seeing any SSE instructions here?) but whatever it is, it's being called
from the IR I attached earlier, and is modifying ECX under some
circumstances.
On 19/07/2013 3:29 PM, Craig Topper wrote:
> That should map directly to sqrtpd which can't modif...
2013 Jul 19
2
[LLVMdev] SIMD instructions and memory alignment on X86
That should map directly to sqrtpd which can't modify ecx.
On Thu, Jul 18, 2013 at 10:27 PM, Peter Newman <peter at uformia.com> wrote:
> Sorry, that should have been llvm.x86.sse2.sqrt.pd
>
>
> On 19/07/2013 3:25 PM, Craig Topper wrote:
>
> What is "frep.x86.sse2.sqrt.pd". I'm only fami...
2013 Jul 19
0
[LLVMdev] SIMD instructions and memory alignment on X86
Sorry, that should have been llvm.x86.sse2.sqrt.pd
On 19/07/2013 3:25 PM, Craig Topper wrote:
> What is "frep.x86.sse2.sqrt.pd". I'm only familiar with things
> prefixed with "llvm.x86".
>
>
> On Thu, Jul 18, 2013 at 10:12 PM, Peter Newman <peter at uformia.com
> <mailto:peter at uformia.com>> wrote:
>
> After stepping through the
2013 Jul 19
2
[LLVMdev] SIMD instructions and memory alignment on X86
What is "frep.x86.sse2.sqrt.pd". I'm only familiar with things prefixed
with "llvm.x86".
On Thu, Jul 18, 2013 at 10:12 PM, Peter Newman <peter at uformia.com> wrote:
> After stepping through the produced assembly, I believe I have a culprit.
>
> One of the calls to @frep.x86.sse2.sqrt.pd is modifying the value of ECX -
> while the produced code is