<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">
<html lang="en">
<head>
<meta content="text/html; charset=US-ASCII" http-equiv="Content-Type">
<title>
GitLab
</title>



<style>img {
max-width: 100%; height: auto;
}
</style>
</head>
<body>
<div class="content">

<h3>
 Marge Bot pushed to branch wip/marge_bot_batch_merge_job
at <a href="https://gitlab.haskell.org/ghc/ghc">Glasgow Haskell Compiler / GHC</a>
</h3>
<h4>
Commits:
</h4>
<ul>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/2485c08aeecd9664bca7812e69e59b85ce16c079">2485c08a</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Skip dynamicToo006 when dynamic linking is not available

This was previously failling on Windows.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/aa5164313aa86941bf15722936824dfbf556a911">aa516431</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark T3372 as fragile on Windows

On Windows we must lock package databases even when opening for
read-only access. This means that concurrent GHC sessions are very
likely to fail with file lock contention.

See #16773.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/2eedb120a6fae214ca5b55658939a359bc55d19a">2eedb120</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Add stderr output for UnsafeInfered02 on Windows

This test uses TemplateHaskell causing GHC to build dynamic objects on
platforms where dynamic linking is available. However, Windows doesn't support
dynamic linking. Consequently the test would fail on Windows with:

```patch
--- safeHaskell/safeInfered/UnsafeInfered02.run/UnsafeInfered02.stderr.normalised       2019-06-04 15:10:10.521594200 +0000
+++ safeHaskell/safeInfered/UnsafeInfered02.run/UnsafeInfered02.comp.stderr.normalised  2019-06-04 15:10:10.523546200 +0000
@@ -1,5 +1,5 @@
-[1 of 2] Compiling UnsafeInfered02_A ( UnsafeInfered02_A.hs, UnsafeInfered02_A.o, UnsafeInfered02_A.dyn_o )
-[2 of 2] Compiling UnsafeInfered02  ( UnsafeInfered02.hs, UnsafeInfered02.o, UnsafeInfered02.dyn_o )
+[1 of 2] Compiling UnsafeInfered02_A ( UnsafeInfered02_A.hs, UnsafeInfered02_A.o )
+[2 of 2] Compiling UnsafeInfered02  ( UnsafeInfered02.hs, UnsafeInfered02.o )

 UnsafeInfered02.hs:4:1:
     UnsafeInfered02_A: Can't be safely imported!
```

The other approach I considered for this issue is to pass `-v0` to GHC.
However, I felt we should probably do this consistently for all of the tests in
this directory and this would take more time than I currently have.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/3967d13a2037d65d4f00fc43076b53339b475876">3967d13a</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark OldModLocation as broken on Windows

Strangely the path it emits contains duplicate path delimiters (#16772),
```patch
--- ghc-api/downsweep/OldModLocation.run/OldModLocation.stderr.normalised       2019-06-04 14:40:26.326075000 +0000
+++ ghc-api/downsweep/OldModLocation.run/OldModLocation.run.stderr.normalised   2019-06-04 14:40:26.328029200 +0000
@@ -1 +1 @@
-[Just "A.hs",Just "mydir/B.hs"]
+[Just "A.hs",Just "mydir//B.hs"]
```
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/31f2ea6899041974c601ca368b9c2af5e8de125b">31f2ea68</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark T7170 as broken on Windows

Due to #16801.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/7bd1c3e15f83c04447bc2ec41bbc36c9c1b7fe0d">7bd1c3e1</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark T7702 as broken on Windows

Due to #16799.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/84900724d59f092a6eebe2834c1c5a62186e430e">84900724</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark T15633a and T15633b as fragile on Windows

As noted in #16813, these tests seem to be fragile on Windows.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/49fff41d0385ac1ad80a1a8b2ae33e765e89a088">49fff41d</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-21T17:32:34Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">linker: Disable code unloading

As noted in #16841, there are currently a variety of bugs in the
unloading logic. These only affect Windows since code unloading is
disabled on Linux, where we build with `GhcDynamic=YES` by default.

In the interest of getting the tree green on Windows disable code
unloading until the issues are resolved.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/e0595d22ce5bc19699079abdb47377b5707cdbbc">e0595d22</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-22T13:36:53Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark T16608_* as fragile on Darwin

As noted in #16855.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/655c6e265e06acbebcb6f9aa084efb3ce933e189">655c6e26</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-22T14:06:05Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">ghci: Don't rely on resolution of System.IO to base module

Previously we would hackily evaluate a textual code snippet to compute
actions to disable I/O buffering and flush the stdout/stderr handles.
This broke in a number of ways (#15336, #16563).

Instead we now ship a module (`GHC.GHCi.Helpers`) with `base` containing
the needed actions. We can then easily refer to these via `Orig` names.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/8f8fc31b84919ea40da87eb4607bb2a52b960efe">8f8fc31b</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-22T14:06:05Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Add test for #16563
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/22e721c10e3790dac6113cdcb33c62c07db54cb0">22e721c1</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-22T14:06:05Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Mark T5611 as broken in ghci way

As described in #16845.
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/b0d6bf2a34d5e2e0cfb2410f58e768935d527be0">b0d6bf2a</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-22T14:06:05Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">rts: Reset STATIC_LINK field of reverted CAFs

When we revert a CAF we must reset the STATIC_LINK field lest the GC
might ignore the CAF (e.g. as it carries the STATIC_FLAG_LIST flag) and
will consequently overlook references to object code that we are trying
to unload. This would result in the reachable object code being
unloaded. See Note [CAF lists] and Note [STATIC_LINK fields].

This fixes #16842.

Idea-due-to: Phuong Trinh <lolotp@fb.com>
</pre>
</li>
<li>
<strong><a href="https://gitlab.haskell.org/ghc/ghc/commit/1f2fff89afebb065b27eba0f6e1f89e25c1c158d">1f2fff89</a></strong>
<div>
<span>by Ben Gamari</span>
<i>at 2019-06-22T14:06:05Z</i>
</div>
<pre class="commit-message" style="white-space: pre-wrap; margin: 0;">testsuite: Add caf_crash testcase
</pre>
</li>
</ul>
<h4>30 changed files:</h4>
<ul>
<li class="file-stats">
<a href="#47f4bea1d308dee249030e6063d92071b5aac240">
compiler/ghci/Linker.hs
</a>
</li>
<li class="file-stats">
<a href="#7b2a3c8a780cfa42412ee2700454599d02c3f69d">
compiler/prelude/PrelNames.hs
</a>
</li>
<li class="file-stats">
<a href="#57d22eb49f269b51f07e23c509834d85e25f691e">
ghc/GHCi/UI/Monad.hs
</a>
</li>
<li class="file-stats">
<a href="#d323d108fe74db3c82576e0a81b4edc5c07a5bfc">
<span class="new-file">
+
libraries/base/GHC/GHCi/Helpers.hs
</span>
</a>
</li>
<li class="file-stats">
<a href="#b7cfb47853ef95330e499ca0c5b90964744d101f">
libraries/base/base.cabal
</a>
</li>
<li class="file-stats">
<a href="#7e3bf511f901d9696a9dc8fa89ada599424b1a43">
libraries/ghc-boot/GHC/PackageDb.hs
</a>
</li>
<li class="file-stats">
<a href="#a481abfdf0106c53321542436ad78f6df542b1b8">
rts/sm/GCAux.c
</a>
</li>
<li class="file-stats">
<a href="#40142624dc4857a702b52f173a35277f48e3a483">
testsuite/tests/concurrent/should_run/all.T
</a>
</li>
<li class="file-stats">
<a href="#f4d80885c91f79726c91193d88da9926499de467">
testsuite/tests/driver/T16608/all.T
</a>
</li>
<li class="file-stats">
<a href="#bd582795756437dfe5c8daa9a9bf24ecdf094888">
testsuite/tests/driver/dynamicToo/dynamicToo006/all.T
</a>
</li>
<li class="file-stats">
<a href="#fbea9a6392407b0fa9ede5b9aa0bb2445af9de4b">
testsuite/tests/ffi/should_run/all.T
</a>
</li>
<li class="file-stats">
<a href="#4ced41462cd902cc48ac9b0ccb60f55a34b517bb">
testsuite/tests/ghc-api/downsweep/all.T
</a>
</li>
<li class="file-stats">
<a href="#8fba858c8f3f59be9a9af7fc941786ece8ec2410">
testsuite/tests/ghci.debugger/scripts/break006.stdout
</a>
</li>
<li class="file-stats">
<a href="#dbebcb81c3d1384f84270821f8eb4023e29522db">
testsuite/tests/ghci.debugger/scripts/break013.stdout
</a>
</li>
<li class="file-stats">
<a href="#31879efd41705720c2f093d77c4156faecb253fd">
testsuite/tests/ghci.debugger/scripts/hist001.stdout
</a>
</li>
<li class="file-stats">
<a href="#62edb1729b9cb606143d1a7f0181f03c3a6b2bda">
testsuite/tests/ghci.debugger/scripts/hist002.stdout
</a>
</li>
<li class="file-stats">
<a href="#a127a665824482697556e32d3e501cdbd9f50404">
<span class="new-file">
+
testsuite/tests/ghci/caf_crash/A.hs
</span>
</a>
</li>
<li class="file-stats">
<a href="#c884e90f81f6a2aae3a08df5dc0d1329d2cafe4a">
<span class="new-file">
+
testsuite/tests/ghci/caf_crash/B.hs
</span>
</a>
</li>
<li class="file-stats">
<a href="#8dd7e9b031dcbb4238cf55a77b490761e402ef6a">
<span class="new-file">
+
testsuite/tests/ghci/caf_crash/D.hs
</span>
</a>
</li>
<li class="file-stats">
<a href="#032a110aa4888251675650dc78c87a79f0dcb32a">
<span class="new-file">
+
testsuite/tests/ghci/caf_crash/all.T
</span>
</a>
</li>
<li class="file-stats">
<a href="#0c521939dc51990274d2f0740d45d4865aa17522">
<span class="new-file">
+
testsuite/tests/ghci/caf_crash/caf_crash.script
</span>
</a>
</li>
<li class="file-stats">
<a href="#886ef18edd4ec2e3ead03d76473d22307f18b152">
<span class="new-file">
+
testsuite/tests/ghci/caf_crash/caf_crash.stdout
</span>
</a>
</li>
<li class="file-stats">
<a href="#7d4e0f1430429d72ad482b06f54174de16ff1c1f">
testsuite/tests/ghci/linking/dyn/all.T
</a>
</li>
<li class="file-stats">
<a href="#a2173f1fb1a762ebc0bda4da3e64d890d1623693">
<span class="new-file">
+
testsuite/tests/ghci/scripts/T16563.script
</span>
</a>
</li>
<li class="file-stats">
<a href="#c710e41aa5dbd8e355c0e903f8cb86b93e16af57">
<span class="new-file">
+
testsuite/tests/ghci/scripts/T16563.stdout
</span>
</a>
</li>
<li class="file-stats">
<a href="#8da75825433bf887a762deedc637b1c15b637d5d">
testsuite/tests/ghci/scripts/T4175.stdout
</a>
</li>
<li class="file-stats">
<a href="#35c9dead622e26b830535877ec549257bcaf1ac1">
testsuite/tests/ghci/scripts/T7627.stdout
</a>
</li>
<li class="file-stats">
<a href="#f0838d8eb1a4af059d02280d76b2a0c5179d5a5a">
testsuite/tests/ghci/scripts/T8469.stdout
</a>
</li>
<li class="file-stats">
<a href="#080fb38cc741d3f1a476696b3373ab7cf48c2aaa">
testsuite/tests/ghci/scripts/all.T
</a>
</li>
<li class="file-stats">
<a href="#aa84ef307ee23e395c25ee686d2a07a07d2e3b5a">
testsuite/tests/ghci/scripts/ghci011.stdout
</a>
</li>
</ul>
<h5>The diff was not included because it is too large.</h5>

</div>
<div class="footer" style="margin-top: 10px;">
<p style="font-size: small; color: #777;">

<br>
<a href="https://gitlab.haskell.org/ghc/ghc/compare/1305202020ab3a306dd66bb921731226fefcf722...1f2fff89afebb065b27eba0f6e1f89e25c1c158d">View it on GitLab</a>.
<br>
You're receiving this email because of your account on gitlab.haskell.org.
If you'd like to receive fewer emails, you can
adjust your notification settings.


</p>
</div>
</body>
</html>