Page MenuHomePhabricator

lebedev.ri (Roman Lebedev)
User

Projects

User does not belong to any projects.

User Details

User Since
Oct 27 2012, 6:35 AM (406 w, 4 d)

Recent Activity

Today

lebedev.ri resigned from D85583: [MemDep] Use BatchAA when computing pointer dependencies.

I applaud the change, but i'm probably not the right reviewer for this.

Wed, Aug 12, 2:59 PM · Restricted Project
lebedev.ri added a comment to D85787: [InstCombine] Aggregate reconstruction simplification (PR47060).

@ reviewers - i'm not so much interested in deep code/algo review,
but more like in the general direction disscussion, like, is this okay for instcombine? :)

Wed, Aug 12, 2:40 PM · Restricted Project
lebedev.ri resigned from D81833: [X86][FPEnv] Lowering of {get,set,reset}_fpenv.
Wed, Aug 12, 2:36 PM · Restricted Project
lebedev.ri resigned from D85528: [analyzer] Fix cast evaluation on scoped enums in ExprEngine.
Wed, Aug 12, 2:34 PM · Restricted Project
lebedev.ri requested changes to D75980: [LV] Generate RT checks up-front and remove them if required. (WIP).

(marking as reviewed)

Wed, Aug 12, 2:33 PM · Restricted Project
lebedev.ri abandoned D75505: [InstCombine] Enhance cast-of-load->load-of-bitcast fold to handle multiple identical casts.
Wed, Aug 12, 2:26 PM · Restricted Project
lebedev.ri abandoned D51570: [X86][XOP] VFRCZ* instructions should be in their own sched class.
Wed, Aug 12, 2:26 PM
lebedev.ri abandoned D69387: [ConstantRange] Add toKnownBits() method.
Wed, Aug 12, 2:24 PM · Restricted Project
lebedev.ri abandoned D75510: [X86][llvm-exegesis] Exploring vector insert/extract.
Wed, Aug 12, 2:23 PM · Restricted Project
lebedev.ri planned changes to D84609: [MemDepAnalysis] Cut-off threshold reshuffling .

As per @asbirlea's comments.

Wed, Aug 12, 2:23 PM · Restricted Project
lebedev.ri planned changes to D84742: [NFCI]MemDepAnalysis] Introduce global limit on a number of instructions to be traversed during single query.

Not needed unless D84609 proceeds.

Wed, Aug 12, 2:23 PM · Restricted Project
lebedev.ri committed rGd6f0600c96a6: [NFC][InstCombine] Add FIXME's for getLogBase2() / visitUDivOperand() (authored by lebedev.ri).
[NFC][InstCombine] Add FIXME's for getLogBase2() / visitUDivOperand()
Wed, Aug 12, 12:07 PM
lebedev.ri committed rG12d93a27e7b7: [InstCombine] Sanitize undef vector constant to 1 in X*(2^C) with X << C… (authored by lebedev.ri).
[InstCombine] Sanitize undef vector constant to 1 in X*(2^C) with X << C…
Wed, Aug 12, 12:07 PM
lebedev.ri requested changes to D76434: [SCEV] Query expanded immediate cost at minsize.

Apologies if i'm missing the point here.

Wed, Aug 12, 1:50 AM · Restricted Project
lebedev.ri updated the diff for D85787: [InstCombine] Aggregate reconstruction simplification (PR47060).

One last tiny cleanup.

Wed, Aug 12, 12:03 AM · Restricted Project

Yesterday

lebedev.ri added a comment to D82847: [CMAKE] Fix 'clean' target not working.

Reverting this patch seems to fix the error for me. Are you still seeing this same error in trunk?

Tue, Aug 11, 2:58 PM · Restricted Project
lebedev.ri updated the summary of D85787: [InstCombine] Aggregate reconstruction simplification (PR47060).
Tue, Aug 11, 2:55 PM · Restricted Project
lebedev.ri added a comment to D82847: [CMAKE] Fix 'clean' target not working.

This commit appears to be causing an error when I run CMake:

CMake Error:

Running

 '/usr/bin/ninja-build' '-C' '/home/tstellar/llvm-project/llvm-build' '-t' 'cleandead'

failed with:

 ninja: error: remove(include/llvm/Support): Directory not empty

CMake Generate step failed. Build files cannot be regenerated correctly.

Tue, Aug 11, 2:52 PM · Restricted Project
lebedev.ri added a comment to D85647: [InstCombine] eliminate a pointer cast around insertelement.

LGTM, thank you!

Tue, Aug 11, 2:39 PM · Restricted Project
lebedev.ri updated the summary of D85787: [InstCombine] Aggregate reconstruction simplification (PR47060).
Tue, Aug 11, 2:34 PM · Restricted Project
lebedev.ri requested review of D85787: [InstCombine] Aggregate reconstruction simplification (PR47060).
Tue, Aug 11, 2:34 PM · Restricted Project
lebedev.ri committed rGe74e8b2b6957: [NFC][InstCombine] Add tests for PHI merging/aggregate reconstruction (PR47060) (authored by lebedev.ri).
[NFC][InstCombine] Add tests for PHI merging/aggregate reconstruction (PR47060)
Tue, Aug 11, 12:41 PM
lebedev.ri added a comment to D85709: [InstSimplify] Implement Instruction simplification for X/sqrt(X) to sqrt(X)..

I'm fairly sure this transform is a performance loss. For a target like Skylake Server, a SQRT(x) can take up to 20 cycles. But a RSQRT(x) is about 6 cycles and a MUL(y) is 4 cycles. We'd be better off with a X*RSQRT(X).

Tue, Aug 11, 7:27 AM · Restricted Project

Mon, Aug 10

lebedev.ri added a comment to D85639: [Attributor] Fix bug in operator of PotentialValuesState.

Test?

Mon, Aug 10, 4:57 AM · Restricted Project
lebedev.ri added a comment to D73230: [X86][ELF] Prefer to lower MC_GlobalAddress operands to .Lfoo$local.

Passing-by remark: This change passed in our internal huge code base. isn't a great expanded description for the change..

Mon, Aug 10, 4:54 AM · Restricted Project

Sun, Aug 9

lebedev.ri added a comment to D85446: [InstCombine] Add vector support to mul(add(x,c),negpow2) -> mul(sub(-c,x),pow2) folds.

(Just some heads up, in case someone else got the same problem...)

I've noticed some regressions downstream after this patch. Haven't debugged/reduced it completely yet, but at least in one case I suspect that the problem actually might be in load-store-vectorizer.

Used to get something like this before load-store-vectorizer:

<...>

The old input to load-store-vectorizer resulted in two <8 x i16> loads, but the new input gives one <8 x i16 load, two <2 x i16> loads and one <4 x i16> load. So for some reason LSV fails to detect that the loads are consecutive given the new IR. At least that is my current theory.

Sun, Aug 9, 11:12 PM · Restricted Project
lebedev.ri added inline comments to D85593: [InstCombine] ~(~X + Y) -> X - Y.
Sun, Aug 9, 7:23 AM · Restricted Project

Sat, Aug 8

lebedev.ri added a comment to D85593: [InstCombine] ~(~X + Y) -> X - Y.

Let's not introduce any new unneeded one-use checks.

Sat, Aug 8, 11:06 PM · Restricted Project
lebedev.ri committed rGd4c3f20285ad: [Reduce] Rewrite function body delta pass again (authored by lebedev.ri).
[Reduce] Rewrite function body delta pass again
Sat, Aug 8, 1:49 PM
lebedev.ri committed rGc2ebb3246551: [NFC][SimplifyCFG] Add a test showing invoke->call simplification failure (authored by lebedev.ri).
[NFC][SimplifyCFG] Add a test showing invoke->call simplification failure
Sat, Aug 8, 10:05 AM
lebedev.ri committed rGe492f0e03b01: [SimplifyCFG] Fix invoke->call fold w/ multiple invokes in presence of lifetime… (authored by lebedev.ri).
[SimplifyCFG] Fix invoke->call fold w/ multiple invokes in presence of lifetime…
Sat, Aug 8, 10:04 AM
lebedev.ri committed rG1f452ac1d784: [NFC][SimplifyCFG] Rewrite isCleanupBlockEmpty() to be iterator_range-based (authored by lebedev.ri).
[NFC][SimplifyCFG] Rewrite isCleanupBlockEmpty() to be iterator_range-based
Sat, Aug 8, 10:04 AM
lebedev.ri committed rGa587bf3eb074: [NFC][SimplifyCFG] Count the number of invokes turned into calls due to empty… (authored by lebedev.ri).
[NFC][SimplifyCFG] Count the number of invokes turned into calls due to empty…
Sat, Aug 8, 10:04 AM
lebedev.ri accepted D81766: [VectorCombine] try to create vector loads from scalar loads.
Sat, Aug 8, 1:31 AM · Restricted Project

Fri, Aug 7

lebedev.ri added a comment to D75980: [LV] Generate RT checks up-front and remove them if required. (WIP).

I approve of the general direction of this patch series :)
Some high-level notes. I'm not that familiar with the code in question,
so it will be best for someone else to review, too.

Fri, Aug 7, 8:21 AM · Restricted Project
lebedev.ri added a comment to D84399: [SCEVExpander] Avoid re-using existing casts if it means updating users..

@efriedma sudo-reverse-ping, thanks.

Fri, Aug 7, 8:07 AM · Restricted Project
lebedev.ri accepted D85520: [NFC][GVN] Fix "avaliable" typos.

Hm, guess i missed this pattern :/

Fri, Aug 7, 5:50 AM · Restricted Project
lebedev.ri retitled D85520: [NFC][GVN] Fix "avaliable" typos from [GVN] Fix "avaliable" typos to [NFC][GVN] Fix "avaliable" typos.
Fri, Aug 7, 5:50 AM · Restricted Project
lebedev.ri accepted D85189: [SCEV] ] If RHS >= Start, simplify (Start smax RHS) to RHS for trip counts..

I feel like this is pretty identical to the D85046, so lgtm modulo some possible generalization.

Fri, Aug 7, 1:44 AM · Restricted Project
lebedev.ri added a comment to D85504: [Reassociate] [PowerPC] stop common out mul factors if fma is preferred on target.

I'm not sure why i'm added as a reviewer on this patch, i don't work on ppc, i never used it, and i never really worked on that pass.

Fri, Aug 7, 1:36 AM · Restricted Project
lebedev.ri added a comment to D85504: [Reassociate] [PowerPC] stop common out mul factors if fma is preferred on target.

(can you please maybe spellcheck both the patch description and the wording within the patch? it is really hard to read)

Fri, Aug 7, 12:42 AM · Restricted Project

Thu, Aug 6

lebedev.ri committed rGbe02adfad7ac: [InstCombine] Fold (x + C1) * (-1<<C2) --> (-C1 - x) * (1<<C2) (authored by lebedev.ri).
[InstCombine] Fold (x + C1) * (-1<<C2) --> (-C1 - x) * (1<<C2)
Thu, Aug 6, 1:46 PM
lebedev.ri committed rG0c1c756a3153: [InstCombine] Generalize %x * (-1<<C) --> (-%x) * (1<<C) fold (authored by lebedev.ri).
[InstCombine] Generalize %x * (-1<<C) --> (-%x) * (1<<C) fold
Thu, Aug 6, 1:46 PM
lebedev.ri committed rGa404acb86af7: [NFC][InstCombine] Add some more tests for negation sinking into mul (authored by lebedev.ri).
[NFC][InstCombine] Add some more tests for negation sinking into mul
Thu, Aug 6, 1:45 PM
lebedev.ri committed rG7ce76b06ec90: [InstCombine] Fold sdiv exact X, -1<<C --> -(ashr exact X, C) (authored by lebedev.ri).
[InstCombine] Fold sdiv exact X, -1<<C --> -(ashr exact X, C)
Thu, Aug 6, 1:45 PM
lebedev.ri committed rG47aec80e4afc: [NFC][InstCombine] Negator: add a comment about negating exact arithmentic shift (authored by lebedev.ri).
[NFC][InstCombine] Negator: add a comment about negating exact arithmentic shift
Thu, Aug 6, 1:45 PM
lebedev.ri committed rG442cb88f5344: [InstCombine] Generalize sdiv exact X, 1<<C --> ashr exact X, C fold to… (authored by lebedev.ri).
[InstCombine] Generalize sdiv exact X, 1<<C --> ashr exact X, C fold to…
Thu, Aug 6, 1:45 PM
lebedev.ri closed D85446: [InstCombine] Add vector support to mul(add(x,c),negpow2) -> mul(sub(-c,x),pow2) folds.
Thu, Aug 6, 1:45 PM · Restricted Project
lebedev.ri committed rG8633a0d985f1: [NFC][InstCombine] Better tests for x s/EXACT (1 << y) pattern (authored by lebedev.ri).
[NFC][InstCombine] Better tests for x s/EXACT (1 << y) pattern
Thu, Aug 6, 1:45 PM
lebedev.ri committed rG1c21635c94df: [NFC][InstCombine] Tests for x s/EXACT (-1 << y) pattern (authored by lebedev.ri).
[NFC][InstCombine] Tests for x s/EXACT (-1 << y) pattern
Thu, Aug 6, 1:45 PM
lebedev.ri added inline comments to D85446: [InstCombine] Add vector support to mul(add(x,c),negpow2) -> mul(sub(-c,x),pow2) folds.
Thu, Aug 6, 1:13 PM · Restricted Project
lebedev.ri added inline comments to D85446: [InstCombine] Add vector support to mul(add(x,c),negpow2) -> mul(sub(-c,x),pow2) folds.
Thu, Aug 6, 1:01 PM · Restricted Project
lebedev.ri commandeered D85446: [InstCombine] Add vector support to mul(add(x,c),negpow2) -> mul(sub(-c,x),pow2) folds.
Thu, Aug 6, 10:08 AM · Restricted Project
lebedev.ri added a comment to D85446: [InstCombine] Add vector support to mul(add(x,c),negpow2) -> mul(sub(-c,x),pow2) folds.

Aha, good catch.
But this is not enough of a generalization:

diff
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
index 172b304b776..06a0fcbf0fa 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
@@ -233,29 +233,11 @@ Instruction *InstCombinerImpl::visitMul(BinaryOperator &I) {
     }
   }
Thu, Aug 6, 9:46 AM · Restricted Project
lebedev.ri resigned from D85434: BPF: add a SimplifyCFG IR pass during generic Scalar/IPO optimization.

If that's what's works for BPF..

Thu, Aug 6, 8:39 AM · Restricted Project
lebedev.ri requested changes to D67253: clang-misexpect: a standalone tool for verifying the use of __builtin_expect with PGO data.

Abandon this? D66324 landed.

Thu, Aug 6, 6:35 AM · Restricted Project, Restricted Project
lebedev.ri committed rG141357663e67: [InstCombine] (-NSW x) u<= x --> x s<=0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) u<= x --> x s<=0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rGa512c894768b: [NFC][InstCombine] Refactor '(-NSW x) pred x' fold (authored by lebedev.ri).
[NFC][InstCombine] Refactor '(-NSW x) pred x' fold
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG132be1f50277: [InstCombine] (-NSW x) u< x --> x s< 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) u< x --> x s< 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG0e1241a3c988: [InstCombine] (-NSW x) u>= x --> x s>= 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) u>= x --> x s>= 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG16c642fa39d5: [InstCombine] (-NSW x) u> x --> x s> 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) u> x --> x s> 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG01a6c4bd26a5: [InstCombine] (-NSW x) s< x --> x s> 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) s< x --> x s> 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG59387c0dd742: [InstCombine] (-NSW x) s<= x --> x s>= 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) s<= x --> x s>= 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG388520765157: [InstCombine] (-NSW x) s>= x --> x s<= 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) s>= x --> x s<= 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG8878b79cfe99: [InstCombine] (-NSW x) ==/!= x --> x ==/!= 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) ==/!= x --> x ==/!= 0 (PR39480)
Thu, Aug 6, 1:52 AM
lebedev.ri committed rG5060f5682b01: [InstCombine] (-NSW x) s> x --> x s< 0 (PR39480) (authored by lebedev.ri).
[InstCombine] (-NSW x) s> x --> x s< 0 (PR39480)
Thu, Aug 6, 1:51 AM
lebedev.ri committed rG664e1784cd5e: [NFC][InstCombine] Add tests for comparisons between x and negation of x… (authored by lebedev.ri).
[NFC][InstCombine] Add tests for comparisons between x and negation of x…
Thu, Aug 6, 1:51 AM

Wed, Aug 5

lebedev.ri added inline comments to D84947: Add libFuzzer shared object build output.
Wed, Aug 5, 1:13 PM · Restricted Project, Restricted Project
lebedev.ri added inline comments to D84947: Add libFuzzer shared object build output.
Wed, Aug 5, 12:09 PM · Restricted Project, Restricted Project
lebedev.ri committed rGf3056dcc02d5: [InstCombine] Negator: -(cond ? x : -x) --> cond ? -x : x (authored by lebedev.ri).
[InstCombine] Negator: -(cond ? x : -x) --> cond ? -x : x
Wed, Aug 5, 11:48 AM
lebedev.ri committed rG1d25d0734a15: [NFC][InstCombine] Add tests for negation of old-style [n]abs, select-of-op-vs… (authored by lebedev.ri).
[NFC][InstCombine] Add tests for negation of old-style [n]abs, select-of-op-vs…
Wed, Aug 5, 11:48 AM
lebedev.ri committed rGa05ec856a3e7: [NFC][InstCombine] Negator: include all the needed headers, IWYU (authored by lebedev.ri).
[NFC][InstCombine] Negator: include all the needed headers, IWYU
Wed, Aug 5, 10:13 AM
lebedev.ri committed rG3a3c9519e272: [InstCombine] Negator: 0 - (X + Y) --> (-X) - Y iff a single operand negated (authored by lebedev.ri).
[InstCombine] Negator: 0 - (X + Y) --> (-X) - Y iff a single operand negated
Wed, Aug 5, 10:02 AM
lebedev.ri committed rG26f79e258ffd: [NFC][InstCombine] Tests for negation of `add` w/ single negatible operand (authored by lebedev.ri).
[NFC][InstCombine] Tests for negation of `add` w/ single negatible operand
Wed, Aug 5, 10:02 AM
lebedev.ri added a reverting change for rG8aeb2fe13a41: [InstCombine] Negator: -(X << C) --> X * (-1 << C): rGf5df5cd5586a: Recommit "[InstCombine] Negator: -(X << C) --> X * (-1 << C)".
Wed, Aug 5, 5:59 AM
lebedev.ri added a reverting change for rGac70b37a00dc: Revert "[InstCombine] Negator: -(X << C) --> X * (-1 << C)": rGf5df5cd5586a: Recommit "[InstCombine] Negator: -(X << C) --> X * (-1 << C)".
Wed, Aug 5, 5:59 AM
lebedev.ri committed rGf5df5cd5586a: Recommit "[InstCombine] Negator: -(X << C) --> X * (-1 << C)" (authored by lebedev.ri).
Recommit "[InstCombine] Negator: -(X << C) --> X * (-1 << C)"
Wed, Aug 5, 5:59 AM
lebedev.ri retitled D85212: [flang] Add parser support for OpenMP allocate clause from Add parser support for OpenMP allocate clause to [flang] Add parser support for OpenMP allocate clause.
Wed, Aug 5, 5:06 AM · Restricted Project, Restricted Project
lebedev.ri accepted D83216: [Intrinsic] Add sshl.sat/ushl.sat, saturated shift intrinsics..

LGTM, but i would have preferred to see more feedback on the RFC thread.
That being said, unless you are planning on forming calls to these intrinsics in middle-end transform passes
(i.e. only planning on using them in clang codegen), i think this is okay to proceed.

Wed, Aug 5, 5:02 AM · Restricted Project
lebedev.ri added a comment to D84763: DomTree: Make PostDomTree immune to block successors swap.

@kuhar thank you for the review!
@yrouban thank you for the fix!

Wed, Aug 5, 4:05 AM · Restricted Project
lebedev.ri requested changes to D84495: [InstCombine] Disable branch predicate canonicalization by default.

Abandon now that D84763 has resolved the base problem?

Wed, Aug 5, 4:05 AM · Restricted Project
lebedev.ri requested changes to D84493: [InstCombine] Disable branch predicate canonicalization.

Abandon now that D84763 has resolved the base problem?

Wed, Aug 5, 4:05 AM · Restricted Project
lebedev.ri requested changes to D84492: [SimplifyCFG] Enable branch predicate canonicalization by default.

Abandon now that D84763 has resolved the base problem?

Wed, Aug 5, 4:05 AM · Restricted Project
lebedev.ri requested changes to D84491: [SimplifyCFG] Canonicalize branch predicates.

Abandon now that D84763 has resolved the base problem?

Wed, Aug 5, 4:05 AM · Restricted Project
lebedev.ri requested changes to D81089: Mark InstCombine as not preserving CFG.

Glad to see that D84763 is making it's way!

Wed, Aug 5, 4:05 AM · Restricted Project

Tue, Aug 4

lebedev.ri added a reverting change for rG8aeb2fe13a41: [InstCombine] Negator: -(X << C) --> X * (-1 << C): rGac70b37a00dc: Revert "[InstCombine] Negator: -(X << C) --> X * (-1 << C)".
Tue, Aug 4, 5:20 PM
lebedev.ri committed rGac70b37a00dc: Revert "[InstCombine] Negator: -(X << C) --> X * (-1 << C)" (authored by lebedev.ri).
Revert "[InstCombine] Negator: -(X << C) --> X * (-1 << C)"
Tue, Aug 4, 5:20 PM
lebedev.ri committed rG8aeb2fe13a41: [InstCombine] Negator: -(X << C) --> X * (-1 << C) (authored by lebedev.ri).
[InstCombine] Negator: -(X << C) --> X * (-1 << C)
Tue, Aug 4, 5:13 PM
lebedev.ri committed rG8fd57b06a4fd: [NFC][InstCombine] Fix value names (s/%tmp/%i/) and autogenerate a few tests… (authored by lebedev.ri).
[NFC][InstCombine] Fix value names (s/%tmp/%i/) and autogenerate a few tests…
Tue, Aug 4, 5:13 PM
lebedev.ri committed rG3f3303324eb2: [NFC][InstCombine] Negator: add tests for negation of left-shift by constant (authored by lebedev.ri).
[NFC][InstCombine] Negator: add tests for negation of left-shift by constant
Tue, Aug 4, 5:13 PM
lebedev.ri added a comment to D83216: [Intrinsic] Add sshl.sat/ushl.sat, saturated shift intrinsics..

Still no replies to RFC thread :/

Tue, Aug 4, 1:49 PM · Restricted Project
lebedev.ri updated the summary of D83216: [Intrinsic] Add sshl.sat/ushl.sat, saturated shift intrinsics..
Tue, Aug 4, 1:48 PM · Restricted Project
lebedev.ri added a comment to D85189: [SCEV] ] If RHS >= Start, simplify (Start smax RHS) to RHS for trip counts..

Looks good to me.

Tue, Aug 4, 3:01 AM · Restricted Project
lebedev.ri accepted D85188: [JumpThreading] Remove cast's constraint.

lg

Tue, Aug 4, 2:42 AM · Restricted Project

Mon, Aug 3

lebedev.ri added a comment to D84410: [LAA] Dynamically allocate RuntimeCheckingPtrGroup (NFC)..

Ignoring the question of direction, the code change itself seems ok to me.

Mon, Aug 3, 10:34 AM · Restricted Project
lebedev.ri added a comment to D84406: [LoopVersion] Unify SCEVChecks and alias check handling (NFC)..

I'm not really familiar with code in question, but ignoring
the question of whether or not this is moving in the right
direction, the code change itself looks like a cleanup to me.

Mon, Aug 3, 10:31 AM · Restricted Project
lebedev.ri added a comment to D85092: [InstSimplify] Peephole optimization for icmp (urem X, Y), X.

@xldenis if you need help committing do state so, and specify the Author: name <e@ma.il> line for commit attribution.

Mon, Aug 3, 6:08 AM · Restricted Project
lebedev.ri accepted D85130: [SCEVExpander] [PowerPC] clear scev rewriter before deleting instructions..

Not sure this needs review.

Mon, Aug 3, 6:07 AM · Restricted Project
lebedev.ri added a comment to D74436: Change clang option -ffp-model=precise to select ffp-contract=on.

<...>

Mon, Aug 3, 4:57 AM · Restricted Project
lebedev.ri requested changes to D74436: Change clang option -ffp-model=precise to select ffp-contract=on.

IIUC the patch is currently reverted (in 78654e8511cf16d49f6680d782f3771a767ba942), due to ~20 llvm-test-suite failures.

The failures on X86 are listed below. Unfortunately most tests just compare the hash of the output, so it is not obvious in what way the results are changed.

Presumably it means that whatever optimization options those test specify, are now either more strict or less strict.
I think this is really missing some codegen tests that show how codegen actually changes by this.

Mon, Aug 3, 4:23 AM · Restricted Project