Displaying 5 results from an estimated 5 matches for "rl323309".
2018 Jan 25
2
Late setting of SCEV NoWrap flags does bad with cache
...ely, but in fact it takes full traversal through all cached SCEVs at least to identify the ones that depend on the updated AddRec, and this completely ruins the idea of saving compile time.
You may find the example of such behavior if you re-enable the assertion from patch https://reviews.llvm.org/rL323309 . The test that was added with this patch fails on this assert with exactly this problem.
My question is: do we still want to have this hack with late setting of nsw/nuw given that this flag does not give us any benefits in many real situations? What would you think?
Best regards,
Max
----------...
2018 Jan 25
1
Late setting of SCEV NoWrap flags does bad with cache
...ely, but in fact it takes full traversal through all cached SCEVs at least to identify the ones that depend on the updated AddRec, and this completely ruins the idea of saving compile time.
You may find the example of such behavior if you re-enable the assertion from patch https://reviews.llvm.org/rL323309 . The test that was added with this patch fails on this assert with exactly this problem.
My question is: do we still want to have this hack with late setting of nsw/nuw given that this flag does not give us any benefits in many real situations? What would you think?
Best regards,
Max
----------...
2018 Jan 26
0
Late setting of SCEV NoWrap flags does bad with cache
...xploiting the fact that we
don't delete SCEVs so use lists don't need fast deletion. Again,
you'll need to show compile time memory usage does not go up by very
much.
> You may find the example of such behavior if you re-enable the assertion
> from patch https://reviews.llvm.org/rL323309 . The test that was added with
For that case specifically: I don't think such asserts are a good
idea, unless you've already checked SE.isKnownNonNegative(X) (i.e. not
only do you know that X is "supposed to be" non-negative, but you also
know that SCEV can prove it). For insta...
2018 Jan 26
2
Late setting of SCEV NoWrap flags does bad with cache
...ploiting the fact that we don't delete SCEVs so use lists don't need fast deletion. Again, you'll need to show compile time memory usage does not go up by very much.
> You may find the example of such behavior if you re-enable the
> assertion from patch https://reviews.llvm.org/rL323309 . The test that
> was added with
For that case specifically: I don't think such asserts are a good idea, unless you've already checked SE.isKnownNonNegative(X) (i.e. not only do you know that X is "supposed to be" non-negative, but you also know that SCEV can prove it). For...
2018 Jan 26
0
Late setting of SCEV NoWrap flags does bad with cache
...e fact that we don't delete SCEVs so use lists don't need fast deletion. Again, you'll need to show compile time memory usage does not go up by very much.
>
>> You may find the example of such behavior if you re-enable the
>> assertion from patch https://reviews.llvm.org/rL323309 . The test that
>> was added with
>
> For that case specifically: I don't think such asserts are a good idea, unless you've already checked SE.isKnownNonNegative(X) (i.e. not only do you know that X is "supposed to be" non-negative, but you also know that SCEV can pro...