[Issue 18839] New: Crash on getMember of function masking imported template function
https://issues.dlang.org/show_bug.cgi?id=18839 Issue ID: 18839 Summary: Crash on getMember of function masking imported template function Product: D Version: D2 Hardware: x86_64 OS: Linux Status: NEW Severity: normal Priority: P1 Component: dmd Assignee: nob...@puremagic.com Reporter: default_357-l...@yahoo.de bar.d: void fun(T)(); foo.d: module foo; import bar : fun; void fun() { } // either of those lines crashes pragma(msg, __traits(getMember, foo, "fun").stringof); static assert(__traits(compiles, __traits(getMember, foo, "fun"))); --
[Issue 18839] Crash on getMember of function masking imported template function
https://issues.dlang.org/show_bug.cgi?id=18839 --- Comment #1 from FeepingCreature --- Further reduction: void fun2(T)(); alias fun = fun2; void fun() { } pragma(msg, __traits(getMember, foo, "fun").stringof); static assert(__traits(compiles, __traits(getMember, foo, "fun"))); Note that the order of functions - first alias, then mask - is important. --
[Issue 18837] MMFile should have opDollar
https://issues.dlang.org/show_bug.cgi?id=18837 greenify changed: What|Removed |Added Keywords||bootcamp CC||greeen...@gmail.com --
[Issue 18839] Crash on getMember of function masking imported template function
https://issues.dlang.org/show_bug.cgi?id=18839 --- Comment #2 from FeepingCreature --- Correction- in that case, only the pragma(msg) crashes. --
[Issue 18839] Crash on getMember of function masking imported template function
https://issues.dlang.org/show_bug.cgi?id=18839 --- Comment #3 from FeepingCreature --- Double correction- I had a local patch I'd forgotten about. Both still crash. Reduction is valid. (Sorry.) --
[Issue 18840] New: static foreach over an object's tupleof gives wrong information for const variables
https://issues.dlang.org/show_bug.cgi?id=18840 Issue ID: 18840 Summary: static foreach over an object's tupleof gives wrong information for const variables Product: D Version: D2 Hardware: All OS: All Status: NEW Severity: major Priority: P1 Component: dmd Assignee: nob...@puremagic.com Reporter: goober...@gmail.com Created attachment 1691 --> https://issues.dlang.org/attachment.cgi?id=1691&action=edit Simple test struct with a foreach and a static foreach outputting details at compile time I have only tested this with two traits so far, but it does make generating automated bindings/equivalent objects inaccurate. At a minimum, the wrong __traits information gets returned from: * identifier - returns the name of the static foreach iteration variable * parent - returns the parent of said variable rather than the object type Accessing the variables directly from object.tupleof[ index ] works just fine. The attached test.d shows the issue. Current output when compiling that file: foreach pass: foo has parent SomeStruct bar has parent SomeStruct anotherConst has parent SomeStruct notAConst has parent SomeStruct static foreach pass: var has parent main() bar has parent SomeStruct var has parent main() notAConst has parent SomeStruct Expected output when compiling that file: foreach pass: foo has parent SomeStruct bar has parent SomeStruct anotherConst has parent SomeStruct notAConst has parent SomeStruct static foreach pass: foo has parent SomeStruct bar has parent SomeStruct anotherConst has parent SomeStruct notAConst has parent SomeStruct --
[Issue 17546] Cannot call .stringof on a function symbol if it does not have a no-args overload
https://issues.dlang.org/show_bug.cgi?id=17546 Seb changed: What|Removed |Added CC||greensunn...@gmail.com Severity|normal |major --
[Issue 18828] [-betterC] helpless error in object.d
https://issues.dlang.org/show_bug.cgi?id=18828 Seb changed: What|Removed |Added CC||greensunn...@gmail.com --- Comment #1 from Seb --- It would be really helpful if you could post your code or a snippet that reproduces the error. You can use digger for reducing it: https://github.com/CyberShadow/Digger Are you by chance using classes somewhere? --
[Issue 18828] [-betterC] helpless error in object.d
https://issues.dlang.org/show_bug.cgi?id=18828 --- Comment #2 from Vladimir Panteleev --- (In reply to Seb from comment #1) > You can use digger for reducing it: https://github.com/CyberShadow/Digger I think Sebastian meant to direct you to Dustmite: https://github.com/CyberShadow/DustMite --
[Issue 18817] Include debug (symbols) builds of photos/druntime?
https://issues.dlang.org/show_bug.cgi?id=18817 --- Comment #3 from Manu --- Is there a reason not to include one in the windows install bundle? It would complement the standard windows runtime library distribution. --
[Issue 18828] [-betterC] helpless error in object.d
https://issues.dlang.org/show_bug.cgi?id=18828 --- Comment #3 from Xavier Bigand --- Created attachment 1692 --> https://issues.dlang.org/attachment.cgi?id=1692&action=edit source code that produce the error I don't have any class. api_entry.d file is generated, so all functions are similar, only wglChoosePixelFormat and wglDescribePixelFormat are different from others. The dll_entry.d file could be removed if the -betterC worked. forward_initialization.d contains some specific code to override the wglChoosePixelFormat and wglDescribePixelFormat behavior. The archive also contains the dub file. I'll try DustMite later, but actually my code is pretty minimal, except for the generated file, because I need to export all opengl32 symbols to be able to begin tests (else applications will complains about missing APIs). --
[Issue 18841] New: VisualD - Build Phobos Browse Info needs to not be a manual step
https://issues.dlang.org/show_bug.cgi?id=18841 Issue ID: 18841 Summary: VisualD - Build Phobos Browse Info needs to not be a manual step Product: D Version: D2 Hardware: All OS: Windows Status: NEW Severity: enhancement Priority: P1 Component: visuald Assignee: nob...@puremagic.com Reporter: turkey...@gmail.com We need a sensible way to automate this functionality. Presumably, it needs to be executed once each time you update the toolchain (ie, libraries). I often forget to update it when I update DMD, and none of my colleagues even knew the button was there. Also, it's possible that DMD and/or LDC are in use, and they may each have different libraries... which library is the browseinfo built against? I'd suggest it should be the library active for whichever compiler is in use for the current platform|config. --
[Issue 18841] VisualD - Build Phobos Browse Info needs to not be a manual step
https://issues.dlang.org/show_bug.cgi?id=18841 Manu changed: What|Removed |Added Assignee|nob...@puremagic.com|r.sagita...@gmx.de --
[Issue 18493] [betterC] Can't use aggregated type with postblit
https://issues.dlang.org/show_bug.cgi?id=18493 Mike Franklin changed: What|Removed |Added See Also||https://issues.dlang.org/sh ||ow_bug.cgi?id=18828 --
[Issue 18828] [-betterC] helpless error in object.d
https://issues.dlang.org/show_bug.cgi?id=18828 Mike Franklin changed: What|Removed |Added CC||slavo5...@yahoo.com See Also||https://issues.dlang.org/sh ||ow_bug.cgi?id=18493 --- Comment #4 from Mike Franklin --- I tried looking through the attachment, but there's just too much generated code and it also imports other modules which may contain the offending code. I tried doing a `dub build` on the attachment, but was left with the following error: src\api_entry.d(428,8): Error: module `functions` is in file 'opengl_api\functions.d' which cannot be read So I don't have any way to reproduce the problem. A dustmite reduction would be ideal. This may also be a duplicate of issue 18493, depending on where the precise offending code is. --