-
Notifications
You must be signed in to change notification settings - Fork 29.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
deps: v8, cherry-pick 9365d09, aac2f8c, 47d34a3 #25429
Conversation
CC: @schuay |
You also need to update 'v8_embedder_string': '-node.10', in common.gypi. Which we bump with each commit backported. I think but I'm not sure that it should actually be bumped twice in the PR once for each of the commits backported but @hashseed can confirm as he does this regularly. |
62941f8
to
56aaecc
Compare
56aaecc
to
d4577e4
Compare
👋 flagged a couple folks who I saw reviewing v8 backports recently (CC: @BridgeAR, @TimothyGu). This is an attempt to back port some of the work that @hashseed, @schuay, and myself have done in v8 related to test coverage -- the patch didn't apply perfectly, so would love an extra set of eyes to make sure I made reasonable decisions when back porting. |
|
||
void SourceRangeAstVisitor::MaybeRemoveLastContinuationRange( | ||
ZonePtrList<Statement>* statements) { | ||
if (statements == nullptr || statements->is_empty()) return; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
statements
is sometimes a nullptr
, which was leading to a segfault. This does not seem to be possible in the mainline v8; my guess is that a patch that has not been back-ported eliminated the possibility of expr->body()
returning a nullptr
. It felt like a reasonable stop gap (until we land a fully updated v8) to add the check statements == nullptr
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this is the relevant change.
d4577e4
to
28ccf19
Compare
a6d6939
to
8b58ffe
Compare
@hashseed would be great to get your review on this since you made some of the original changes. |
New coverage build with this PR integrated: https://ci.nodejs.org/view/All/job/node-test-commit-linux-coverage-new/3/ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lgtm!
@mhdawson seems like your test build did not start properly. Could you please have another look at it? |
8b58ffe
to
7657903
Compare
Should this be backported to |
@addaleax I think this would be worth back-porting; I honestly wonder if we could get away with landing it on |
@bcoe In that case you or somebody from the V8 team would need to do the work to backport these patches manually (they have merge conflicts on v11.x). |
@addaleax okay, I'll make an effort to do so after my block of meetings frees up later this afternoon; will update this conversation here if it looks like the back-port is possible. |
Original commit message 9365d09: [coverage] Rework continuation counter handling This changes a few bits about how continuation counters are handled. It introduces a new mechanism that allows removal of a continuation range after it has been created. If coverage is enabled, we run a first post-processing pass on the AST immediately after parsing, which removes problematic continuation ranges in two situations: 1. nested continuation counters - only the outermost stays alive. 2. trailing continuation counters within a block-like structure are removed if the containing structure itself has a continuation. [email protected], [email protected], [email protected] Bug: v8:8381, v8:8539 Change-Id: I6bcaea5060d8c481d7bae099f6db9f993cc30ee3 Reviewed-on: https://chromium-review.googlesource.com/c/1339119 Reviewed-by: Yang Guo <[email protected]> Reviewed-by: Leszek Swirski <[email protected]> Reviewed-by: Georg Neis <[email protected]> Commit-Queue: Jakob Gruber <[email protected]> Cr-Commit-Position: refs/heads/master@{#58443} Refs: v8/v8@9365d09 Original commit message aac2f8c: [coverage] Filter out singleton ranges that alias full ranges Block coverage is based on a system of ranges that can either have both a start and end position, or only a start position (so-called singleton ranges). When formatting coverage information, singletons are expanded until the end of the immediate full parent range. E.g. in: {0, 10} // Full range. {5, -1} // Singleton range. the singleton range is expanded to {5, 10}. Singletons are produced mostly for continuation counters that track whether we execute past a specific language construct. Unfortunately, continuation counters can turn up in spots that confuse our post-processing. For example: if (true) { ... block1 ... } else { ... block2 ... } If block1 produces a continuation counter, it could end up with the same start position as the else-branch counter. Since we merge identical blocks, the else-branch could incorrectly end up with an execution count of one. We need to avoid merging such cases. A full range should always take precedence over a singleton range; a singleton range should never expand to completely fill a full range. An additional post-processing pass ensures this. Bug: v8:8237 Change-Id: Idb3ec7b2feddc0585313810b9c8be1e9f4ec64bf Reviewed-on: https://chromium-review.googlesource.com/c/1273095 Reviewed-by: Georg Neis <[email protected]> Reviewed-by: Yang Guo <[email protected]> Commit-Queue: Jakob Gruber <[email protected]> Cr-Commit-Position: refs/heads/master@{#56531} Refs: v8/v8@aac2f8c deps: V8: backport 47d34a3 Original commit message: Revert "[coverage] change block range to avoid ambiguity." This reverts commit 471fef0469d04d7c487f3a08e81f3d77566a2f50. Reason for revert: A more general fix incoming at https://crrev.com/c/1273095. Original change's description: > [coverage] change block range to avoid ambiguity. > > By moving the block range end to left of closing bracket, > we can avoid ambiguity where an open-ended singleton range > could be both interpreted as inside the parent range, or > next to it. > > R=<U+200B>[email protected] > > Bug: v8:8237 > Change-Id: Ibc9412b31efe900b6d8bff0d8fa8c52ddfbf460a > Reviewed-on: https://chromium-review.googlesource.com/1254127 > Reviewed-by: Georg Neis <[email protected]> > Commit-Queue: Yang Guo <[email protected]> > Cr-Commit-Position: refs/heads/master@{#56347} [email protected],[email protected],[email protected] # Not skipping CQ checks because original CL landed > 1 day ago. Bug: v8:8237 Change-Id: I39310cf3c2f06a0d98ff314740aaeefbfffc0834 Reviewed-on: https://chromium-review.googlesource.com/c/1273096 Reviewed-by: Jakob Gruber <[email protected]> Reviewed-by: Toon Verwaest <[email protected]> Reviewed-by: Yang Guo <[email protected]> Commit-Queue: Jakob Gruber <[email protected]> Cr-Commit-Position: refs/heads/master@{#56513} Refs: v8/v8@47d34a3 PR-URL: nodejs#25429 Backport-PR-URL: nodejs#25728 Reviewed-By: Yang Guo <[email protected]> Reviewed-By: Michael Dawson <[email protected]>
Original commit message 9365d09: [coverage] Rework continuation counter handling This changes a few bits about how continuation counters are handled. It introduces a new mechanism that allows removal of a continuation range after it has been created. If coverage is enabled, we run a first post-processing pass on the AST immediately after parsing, which removes problematic continuation ranges in two situations: 1. nested continuation counters - only the outermost stays alive. 2. trailing continuation counters within a block-like structure are removed if the containing structure itself has a continuation. [email protected], [email protected], [email protected] Bug: v8:8381, v8:8539 Change-Id: I6bcaea5060d8c481d7bae099f6db9f993cc30ee3 Reviewed-on: https://chromium-review.googlesource.com/c/1339119 Reviewed-by: Yang Guo <[email protected]> Reviewed-by: Leszek Swirski <[email protected]> Reviewed-by: Georg Neis <[email protected]> Commit-Queue: Jakob Gruber <[email protected]> Cr-Commit-Position: refs/heads/master@{#58443} Refs: v8/v8@9365d09 Original commit message aac2f8c: [coverage] Filter out singleton ranges that alias full ranges Block coverage is based on a system of ranges that can either have both a start and end position, or only a start position (so-called singleton ranges). When formatting coverage information, singletons are expanded until the end of the immediate full parent range. E.g. in: {0, 10} // Full range. {5, -1} // Singleton range. the singleton range is expanded to {5, 10}. Singletons are produced mostly for continuation counters that track whether we execute past a specific language construct. Unfortunately, continuation counters can turn up in spots that confuse our post-processing. For example: if (true) { ... block1 ... } else { ... block2 ... } If block1 produces a continuation counter, it could end up with the same start position as the else-branch counter. Since we merge identical blocks, the else-branch could incorrectly end up with an execution count of one. We need to avoid merging such cases. A full range should always take precedence over a singleton range; a singleton range should never expand to completely fill a full range. An additional post-processing pass ensures this. Bug: v8:8237 Change-Id: Idb3ec7b2feddc0585313810b9c8be1e9f4ec64bf Reviewed-on: https://chromium-review.googlesource.com/c/1273095 Reviewed-by: Georg Neis <[email protected]> Reviewed-by: Yang Guo <[email protected]> Commit-Queue: Jakob Gruber <[email protected]> Cr-Commit-Position: refs/heads/master@{#56531} Refs: v8/v8@aac2f8c deps: V8: backport 47d34a3 Original commit message: Revert "[coverage] change block range to avoid ambiguity." This reverts commit 471fef0469d04d7c487f3a08e81f3d77566a2f50. Reason for revert: A more general fix incoming at https://crrev.com/c/1273095. Original change's description: > [coverage] change block range to avoid ambiguity. > > By moving the block range end to left of closing bracket, > we can avoid ambiguity where an open-ended singleton range > could be both interpreted as inside the parent range, or > next to it. > > R=<U+200B>[email protected] > > Bug: v8:8237 > Change-Id: Ibc9412b31efe900b6d8bff0d8fa8c52ddfbf460a > Reviewed-on: https://chromium-review.googlesource.com/1254127 > Reviewed-by: Georg Neis <[email protected]> > Commit-Queue: Yang Guo <[email protected]> > Cr-Commit-Position: refs/heads/master@{#56347} [email protected],[email protected],[email protected] # Not skipping CQ checks because original CL landed > 1 day ago. Bug: v8:8237 Change-Id: I39310cf3c2f06a0d98ff314740aaeefbfffc0834 Reviewed-on: https://chromium-review.googlesource.com/c/1273096 Reviewed-by: Jakob Gruber <[email protected]> Reviewed-by: Toon Verwaest <[email protected]> Reviewed-by: Yang Guo <[email protected]> Commit-Queue: Jakob Gruber <[email protected]> Cr-Commit-Position: refs/heads/master@{#56513} Refs: v8/v8@47d34a3 PR-URL: #25429 Backport-PR-URL: #25728 Reviewed-By: Yang Guo <[email protected]> Reviewed-By: Michael Dawson <[email protected]> Reviewed-By: Anna Henningsen <[email protected]>
merge patches related to test coverage:
v8/v8@aac2f8c
v8/v8@9365d09
v8/v8@47d34a3
this should correct some of the oddities seen by @mhdawson in #25157
Checklist
make -j4 test
(UNIX), orvcbuild test
(Windows) passes