DMD Patches for three severe wrong-code bugs#1
Merged
37 commits merged intodlang:masterfrom Feb 6, 2011
Merged
Conversation
added 2 commits
January 26, 2011 22:39
1914 Array initialisation from const array yields memory trample 3198 wrong initializer for structs arrays There was an old incorrect D2-only fix for 1914 (the old fix didn't fix all-zero array initializers). This is replaced with a general solution. 3198 is also fixed, by removing an incorrect special case for arrays.
|
Regarding "is this the best way". If you have push access to the official dmd repo (which it looks like you do), then this isn't the best way. Although it does allow us to see only your commits to dmd pretty easily, it requires more work by someone in the official repo (which might be you!) to pull in your repo and merge it in. You're better off cloning the official repo and creating a branch. Then commit your individual changes to that branch. You can then push your local branch to a remote branch here on the official dmd repo. Walter can review the changes and merge them in (or ask you to merge them in). You get the same effect (committing without messing with master), but less work on pulling in your forked repo. |
Member
|
Push access != push permission. Walter pretty much owns all commits, so pull requests are probably exactly right. |
added 5 commits
January 27, 2011 08:10
(patch by Johannes Pfau)
Collaborator
Author
|
I'm a bit surprised that after I made this pull request, my subsequent commits got added to it. |
added 20 commits
January 28, 2011 08:14
Patch by ibuclaw
D2 version. The basic bug causes a segfault. The special treatment with the 'getVarExp' is needed to avoid a stack overflow; this also happens in D1 (the segfault is D2-specific).
Note that we're using HTML4. When we change to HTML5, this test needs to change to the commented-out version.
Patch by ibuclaw and aziz.koeksal
added 8 commits
February 2, 2011 21:24
When a recursive alias declaration is encountered, mark it as type error (not just as an alias to itself of type error).
…n't compile Introduced on 5th Dec, in the fix for bug 5110. From the comments for 5110: "the patch limits STC propagation via Scope's storage class (sc->stc) only to @safe/@trusted/@System. Other STCs such as const and synchronized are 'pulled' by need of each member out of parent AggregateDeclaration." This regression happened because when deducing a function template, these STCs haven't been pulled from the parent yet. Applying the STCs doesn't happen until FuncDeclaration::semantic. This isn't early enough for function template matching. So we need to complete the storage class of 'this' during matching. It applies to const, immutable, synchronized, and shared classes.
Added here because it was broken by the patch for the test included immediately above it. Doesn't need to run (it fails to compile on 2.051).
Patch by rsinfu
I've put this next to the tests for the two closely related bugs.
Govelius
pushed a commit
to Govelius/dmd
that referenced
this pull request
Jul 18, 2011
Patch type-checked-arrays to compile for win32/dmc.
ghost
mentioned this pull request
skl131313
pushed a commit
to skl131313/dmd
that referenced
this pull request
Jan 24, 2017
Fix building for 32mscoff with MARS.
Darredevil
pushed a commit
to Darredevil/dmd
that referenced
this pull request
Mar 28, 2017
Darredevil
pushed a commit
to Darredevil/dmd
that referenced
this pull request
Mar 31, 2017
MartinNowak
added a commit
that referenced
this pull request
May 8, 2017
fixed visual studio build to generate VERSION instead of verstr.h
Darredevil
pushed a commit
to Darredevil/dmd
that referenced
this pull request
May 17, 2017
Merged
kinke
added a commit
to kinke/dmd
that referenced
this pull request
Aug 26, 2023
I guess the recent removal of module constructors in druntime has brought these to light, for a little LDC ASan smoke test, which newly complains with D v2.105, e.g.: > Direct leak of 78624 byte(s) in 126 object(s) allocated from: > #0 0x55606978c40d in malloc /local/mnt/workspace/tmp/final/llvm-project/compiler-rt/lib/asan/asan_malloc_linux.cpp:145:3 > dlang#1 0x5560698050b1 in _D2rt5minfo11ModuleGroup9sortCtorsMFNbAyaZv
RazvanN7
pushed a commit
that referenced
this pull request
Aug 31, 2023
I guess the recent removal of module constructors in druntime has brought these to light, for a little LDC ASan smoke test, which newly complains with D v2.105, e.g.: > Direct leak of 78624 byte(s) in 126 object(s) allocated from: > #0 0x55606978c40d in malloc /local/mnt/workspace/tmp/final/llvm-project/compiler-rt/lib/asan/asan_malloc_linux.cpp:145:3 > #1 0x5560698050b1 in _D2rt5minfo11ModuleGroup9sortCtorsMFNbAyaZv
thewilsonator
pushed a commit
that referenced
this pull request
Sep 11, 2025
Limit the number of platforms that this is done on. A inspection of some libc implementations of fork has identified the main culprits, don't need to apply this to any others. MacOS testsuite also regressed as a result on calling this code, it's not clear why, but the backtrace is: ``` * thread #1, queue = 'com.apple.main-thread', stop reason = EXC_BAD_INSTRUCTION (code=EXC_I386_INVOP, subcode=0x0) * frame #0: 0x00007ff81abe6ee3 libsystem_platform.dylib`_os_unfair_lock_recursive_abort + 23 frame #1: 0x00007ff81abe12da libsystem_platform.dylib`_os_unfair_lock_lock_slow + 247 frame #2: 0x00007ff81abccd44 libsystem_pthread.dylib`_pthread_atfork_prepare_handlers + 48 frame #3: 0x00007ff825dc2705 libSystem.B.dylib`libSystem_atfork_prepare + 25 frame #4: 0x00007ff81aac17e1 libsystem_c.dylib`fork + 24 frame #5: 0x0000000101f730ee test_runner`core.internal.backtrace.dwarf.resolveAddressesWithAtos(Location[]) + 210 ```
This pull request was closed.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I have created my own DMD repository in git@github:donc/dmd.git (so that I don't interfere with the official one). I don't know if this is the best way of doing this.
Three of the oldest and most severe wrong-code bugs:
1899 AA of fixed-length arrays fails to initialize
1914 Array initialisation from const array yields memory trample
3198 wrong initializer for structs arrays
I haven't merged the patches into the D1 branch, but I have added the tests to the test suite.
Let's see how well this works!