[LLVMdev] RFC: PerfGuide for frontend authors

[Philip Reames]

> On Feb 28, 2015, at 3:01 PM, Björn Steinbrink <bsteinbr at gmail.com> wrote:
> 
> [This time without dropping the list, sorry]
> 
> 2015-02-28 23:50 GMT+01:00 Philip Reames <listmail at philipreames.com>:
> 
>>>> On Feb 28, 2015, at 2:30 PM, Björn Steinbrink <bsteinbr at gmail.com> wrote:
>>>> 
>>>>> On 2015.02.28 14:23:02 -0800, Philip Reames wrote:
>>>>> On 02/28/2015 10:04 AM, Björn Steinbrink wrote:
>>>>> Hi,
>>>>> 
>>>>>>> On 2015.02.28 10:53:35 -0600, Hal Finkel wrote:
>>>>>>> ----- Original Message -----
>>>>>>> From: "Philip Reames" <listmail at philipreames.com>
>>>>>>>> 6. Use the lifetime.start/lifetime.end and
>>>>>>>> invariant.start/invariant.end intrinsics where possible
>>>>>>> Do you find these help in practice?  The few experiments I ran were
>>>>>>> neutral at best and harmful in one or two cases.  Do you have
>>>>>>> suggestions on how and when to use them?
>>>>>> Good point, we should be more specific here. My, admittedly limited,
>>>>>> experience with these is that they're most useful when their
>>>>>> properties are not dynamic -- which perhaps means that they
>>>>>> post-dominate the entry, and are applied to allocas in the entry block
>>>>>> -- and the larger the objects in question, the more the potential
>>>>>> stack-space savings, etc.
>>>>> my experience adding support for the lifetime intrinsics to the rust
>>>>> compiler is largely positive (because our code is very stack heavy at
>>>>> the moment), but we still suffer from missed memcpy optimizations.
>>>>> That happens because I made the lifetime regions as small as possible,
>>>>> and sometimes an alloca starts its lifetime too late for the optimization
>>>>> to happen.  My new (but not yet implemented) approach to to "align" the
>>>>> calls to lifetime.start for allocas with overlapping lifetimes unless
>>>>> there's actually a possibility for stack slot sharing.
>>>>> 
>>>>> For example we currently translate:
>>>>> 
>>>>>   let a = [0; 1000000]; // Array of 1000000 zeros
>>>>>   {
>>>>>     let b = a;
>>>>>   }
>>>>>   let c = something;
>>>>> 
>>>>> to roughly this:
>>>>> 
>>>>>   lifetime.start(a)
>>>>>   memset(a, 0, 1000000)
>>>>>   lifetime.start(b)
>>>>>   memcpy(b, a)
>>>>>   lifetime.end(b)
>>>>>   lifetime.start(c)
>>>>>   lifetime.end(c)
>>>>>   lifetime.end(a)
>>>>> 
>>>>> The lifetime.start call for "b" stops the call-slot (I think)
>>>>> optimization from being applied. So instead this should be translated to
>>>>> something like:
>>>>> 
>>>>>   lifetime.start(a)
>>>>>   lifetime.start(b)
>>>>>   memset(a, 0, 1000000)
>>>>>   memcpy(b, a)
>>>>>   lifetime.end(b)
>>>>>   lifetime.start(c)
>>>>>   lifetime.end(c)
>>>>>   lifetime.end(a)
>>>>> 
>>>>> extending the lifetime of "b" because it overlaps with that of "a"
>>>>> anyway. The lifetime of "c" still starts after the end of "b"'s lifetime
>>>>> because there's actually a possibility for stack slot sharing.
>>>>> 
>>>>> Björn
>>>> I'd be interested in seeing the IR for this that you're currently
>>>> generating.  Unless I'm misreading your example, everything in this is
>>>> completely dead.  We should be able to reduce this to nothing and if we
>>>> can't, it's clearly a missed optimization.  I'm particularly interested in
>>>> how the difference in placement of the lifetime start for 'b' effects
>>>> optimization.  I really wouldn't expect that.
>>> 
>>> I should have clarified that that was a reduced, incomplete example, the
>>> actual code looks like this (after optimizations):
>>> 
>>> define void @_ZN9test_func20hdd8a534ccbedd903paaE(i1 zeroext) unnamed_addr #0 {
>>> entry-block:
>>>   %x = alloca [100000 x i32], align 4
>>>   %1 = bitcast [100000 x i32]* %x to i8*
>>>   %arg = alloca [100000 x i32], align 4
>>>   call void @llvm.lifetime.start(i64 400000, i8* %1)
>>>   call void @llvm.memset.p0i8.i64(i8* %1, i8 0, i64 400000, i32 4, i1 false)
>>>   %2 = bitcast [100000 x i32]* %arg to i8*
>>>   call void @llvm.lifetime.start(i64 400000, i8* %2) ; this happens too late
>>>   call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %1, i64 400000, i32 4, i1 false)
>>>   call void asm "", "r,~{dirflag},~{fpsr},~{flags}"([100000 x i32]* %arg) #2, !noalias !0, !srcloc !3
>>>   call void @llvm.lifetime.end(i64 400000, i8* %2) #2, !alias.scope !4, !noalias !0
>>>   call void @llvm.lifetime.end(i64 400000, i8* %2)
>>>   call void @llvm.lifetime.end(i64 400000, i8* %1)
>>>   ret void
>>> }
>>> 
>>> If the lifetime start for %arg is moved up, before the memset, the
>>> callslot optimization can take place and the %c alloca is eliminated,
>>> but with the lifetime starting after the memset, that isn't possible.
>> This bit of ir actually seems pretty reasonable given the inline asm.  The only thing I really see is that the memcpy could be a memset.  Are you expecting something else?
> 
> The only thing that is to be improved is that the memset should
> directly write to %arg and %x should be removed because it is dead
> then. This happens when there are no lifetime intrinsics or when the
> call to lifetime.start is moved before the call to memset. The latter
> is what my first mail was about, that it is usually better to have
> overlapping lifetimes all start at the same point, instead of starting
> them as late as possible.
Honestly, this sounds like a clear optimizer bug, not something a frontend should work around.

Can you file a bug with the four sets of ir?  (Both schedules, no intrinsics before and after). This should hopefully be easy to fix.  

Do you know of other cases like this with the lifetime intrinsics?

Philip