llvm dev - Oct 2017 - RFC: Generate plain !tbaa tags in place of !tbaa.struct ones

Hello,

The motivation behind this proposal is to make it possible to
propagate TBAA information through scalarizing transformations,
such as SROA, that rewrite accesses to aggregates, e.g., memcpy()
calls, into accesses to scalars, that is, load and store
instructions.

Currently, we decorate instructions that initialize and copy
aggregates with !tbaa.struct tags that generally cannot be
transformed to !tbaa tags. For this reason every time we
scalarize an aggregate, we leave resulting loads and stores
undecorated, meaning optimization of such instructions cannot
benefit from TBAA.

Furthermore, our analysis indicates that the only place where
!tbaa.struct tags may potentially impact code generation is
simplification of memcpy() and memmove() calls, see
SimplifyMemTransfer() in InstCombineCalls.cpp. Ironically, what
the code that makes that sole use of such tags is trying to do is
to construct a !tbaa tag from the information encoded in the
given !tbaa.struct tag. Note that it can only do that if the
!tbaa.struct tag describes a structure with a single member of a
scalar type.

Here's how we propose to resolve the issue in terms of specific
steps:

1. Extend the TBAA facilities in clang to support aggregate types
    as final access types. This patch:

    [CodeGen] Propagate may-alias'ness of lvalues with TBAA info
    https://reviews.llvm.org/D39008

    implements this for the needs of fixing issues with
    propagation of TBAA information, which in turn is necessary to
    support TBAA for unions. So once this patch is committed, this
    step is done.

2. Generate !tbaa tags in addition to !tbaa.struct tags for
    aggregate accesses.

3. Fix the TBAA analysis to support aggregate accesses as
    explained in this proposal:

    RFC: Resolving TBAA issues
    http://lists.llvm.org/pipermail/llvm-dev/2017-August/116652.html

4. Switch the code that simplifies memcpy() and memmove() calls
    to !tbaa tags.

5. Remove the support for !tbaa.struct tags.

    I guess we would want to remove all code which only purpose is
    to generate such tags, but I'm not sure what we should do with
    the MD_tbaa_struct enumerator. Some possible options:
    a) Leave it as is, just don't use it.
    b) Rename to something like MD_unused.
    c) Remove, but do not change the values of other MD_*
       enumerators.
    d) Remove and adjust values of other MD_* enumerators
       respectively.
    Or, maybe we want some multi-stage plan here?

Further steps are supposed to include things like fixing SROA to
propagate TBAA information.

Any feedback is highly appreciated.

Thanks,

--

On 10/18/2017 05:49 AM, Ivan Kosarev via llvm-dev wrote:
> Hello,
>
> The motivation behind this proposal is to make it possible to
> propagate TBAA information through scalarizing transformations,
> such as SROA, that rewrite accesses to aggregates, e.g., memcpy()
> calls, into accesses to scalars, that is, load and store
> instructions.
>
> Currently, we decorate instructions that initialize and copy
> aggregates with !tbaa.struct tags that generally cannot be
> transformed to !tbaa tags. For this reason every time we
> scalarize an aggregate, we leave resulting loads and stores
> undecorated, meaning optimization of such instructions cannot
> benefit from TBAA.
I understand that we'd like to improve TBAA in a number of different 
ways, and that probably involves replacing !tbaa.struct with an enhanced 
!tbaa representation, but I don't understand the point of this proposal. 
As you state, there are lots of places that could use !tbaa.struct 
information, such as SROA, but don't. Is the current !tbaa.struct 
missing some information that is essential for that purpose?

  -Hal
>
> Furthermore, our analysis indicates that the only place where
> !tbaa.struct tags may potentially impact code generation is
> simplification of memcpy() and memmove() calls, see
> SimplifyMemTransfer() in InstCombineCalls.cpp. Ironically, what
> the code that makes that sole use of such tags is trying to do is
> to construct a !tbaa tag from the information encoded in the
> given !tbaa.struct tag. Note that it can only do that if the
> !tbaa.struct tag describes a structure with a single member of a
> scalar type.
>
> Here's how we propose to resolve the issue in terms of specific
> steps:
>
> 1. Extend the TBAA facilities in clang to support aggregate types
>    as final access types. This patch:
>
>    [CodeGen] Propagate may-alias'ness of lvalues with TBAA info
>    https://reviews.llvm.org/D39008
>
>    implements this for the needs of fixing issues with
>    propagation of TBAA information, which in turn is necessary to
>    support TBAA for unions. So once this patch is committed, this
>    step is done.
>
> 2. Generate !tbaa tags in addition to !tbaa.struct tags for
>    aggregate accesses.
>
> 3. Fix the TBAA analysis to support aggregate accesses as
>    explained in this proposal:
>
>    RFC: Resolving TBAA issues
> http://lists.llvm.org/pipermail/llvm-dev/2017-August/116652.html
>
> 4. Switch the code that simplifies memcpy() and memmove() calls
>    to !tbaa tags.
>
> 5. Remove the support for !tbaa.struct tags.
>
>    I guess we would want to remove all code which only purpose is
>    to generate such tags, but I'm not sure what we should do with
>    the MD_tbaa_struct enumerator. Some possible options:
>    a) Leave it as is, just don't use it.
>    b) Rename to something like MD_unused.
>    c) Remove, but do not change the values of other MD_*
>       enumerators.
>    d) Remove and adjust values of other MD_* enumerators
>       respectively.
>    Or, maybe we want some multi-stage plan here?
>
> Further steps are supposed to include things like fixing SROA to
> propagate TBAA information.
>
> Any feedback is highly appreciated.
>
> Thanks,
>
-- 
Hal Finkel
Lead, Compiler Technology and Programming Languages
Leadership Computing Facility
Argonne National Laboratory

In short, the problem with !tbaa.struct is that in most cases it cannot 
be converted to !tbaa. For transformations like SROA this means they 
cannot propagate !tbaa.struct tags for aggregate-accessing instructions 
like memcpy() calls to the resulting loads and stores.


On 31/10/17 05:11, Hal Finkel wrote:
>
> On 10/18/2017 05:49 AM, Ivan Kosarev via llvm-dev wrote:
>> Hello,
>>
>> The motivation behind this proposal is to make it possible to
>> propagate TBAA information through scalarizing transformations,
>> such as SROA, that rewrite accesses to aggregates, e.g., memcpy()
>> calls, into accesses to scalars, that is, load and store
>> instructions.
>>
>> Currently, we decorate instructions that initialize and copy
>> aggregates with !tbaa.struct tags that generally cannot be
>> transformed to !tbaa tags. For this reason every time we
>> scalarize an aggregate, we leave resulting loads and stores
>> undecorated, meaning optimization of such instructions cannot
>> benefit from TBAA.
>
> I understand that we'd like to improve TBAA in a number of different 
> ways, and that probably involves replacing !tbaa.struct with an 
> enhanced !tbaa representation, but I don't understand the point of 
> this proposal. As you state, there are lots of places that could use 
> !tbaa.struct information, such as SROA, but don't. Is the current 
> !tbaa.struct missing some information that is essential for that purpose?
>
>  -Hal
>
>>
>> Furthermore, our analysis indicates that the only place where
>> !tbaa.struct tags may potentially impact code generation is
>> simplification of memcpy() and memmove() calls, see
>> SimplifyMemTransfer() in InstCombineCalls.cpp. Ironically, what
>> the code that makes that sole use of such tags is trying to do is
>> to construct a !tbaa tag from the information encoded in the
>> given !tbaa.struct tag. Note that it can only do that if the
>> !tbaa.struct tag describes a structure with a single member of a
>> scalar type.
>>
>> Here's how we propose to resolve the issue in terms of specific
>> steps:
>>
>> 1. Extend the TBAA facilities in clang to support aggregate types
>>    as final access types. This patch:
>>
>>    [CodeGen] Propagate may-alias'ness of lvalues with TBAA info
>>    https://reviews.llvm.org/D39008
>>
>>    implements this for the needs of fixing issues with
>>    propagation of TBAA information, which in turn is necessary to
>>    support TBAA for unions. So once this patch is committed, this
>>    step is done.
>>
>> 2. Generate !tbaa tags in addition to !tbaa.struct tags for
>>    aggregate accesses.
>>
>> 3. Fix the TBAA analysis to support aggregate accesses as
>>    explained in this proposal:
>>
>>    RFC: Resolving TBAA issues
>> http://lists.llvm.org/pipermail/llvm-dev/2017-August/116652.html
>>
>> 4. Switch the code that simplifies memcpy() and memmove() calls
>>    to !tbaa tags.
>>
>> 5. Remove the support for !tbaa.struct tags.
>>
>>    I guess we would want to remove all code which only purpose is
>>    to generate such tags, but I'm not sure what we should do with
>>    the MD_tbaa_struct enumerator. Some possible options:
>>    a) Leave it as is, just don't use it.
>>    b) Rename to something like MD_unused.
>>    c) Remove, but do not change the values of other MD_*
>>       enumerators.
>>    d) Remove and adjust values of other MD_* enumerators
>>       respectively.
>>    Or, maybe we want some multi-stage plan here?
>>
>> Further steps are supposed to include things like fixing SROA to
>> propagate TBAA information.
>>
>> Any feedback is highly appreciated.
>>
>> Thanks,
>>
>