Re: Step into in debugging.
hmm.. I didn't change my nickname. It was my answer.
Re: Step into in debugging.
Of course no. When I have some problem I google a solution and often I find it. I'm sure that sharing your solution may be useful for somebody else. But very often people that have solved theirs problem do not share solution - they think something like who cares and so on. You didn't shy to publish your problem and solution and I appreciated it, sincerely.
Re: Step into in debugging.
On Wednesday, 22 February 2017 at 14:35:14 UTC, drug wrote: Thank you for sharing this! Sarcasm?
Re: Step into in debugging.
22.02.2017 17:32, ANtlord пишет: On Wednesday, 22 February 2017 at 14:23:02 UTC, ANtlord wrote: My bad again. It appears I have a smaller knowledge that I think. I'm ended up in typecons because `allocator` is made by template `scoped` returns instance of internal static struct `Scoped` and it has alias is asigned to method `Scoped_payload` and this is called on https://github.com/ANtlord/DCD/blob/master/src/server/autocomplete.d#L341. SOLVED And one more. Method `Scoped_payload` returns instance of type passed as template argument for template method `scoped`. As long as method `generateAutocompleteTrees` requires second argument must be instance of `IAllocator`, object `allocator` "transforms" to instance of class implements `IAllocator` by alias of `Scoped_payload` assigned to `this`. Thank you for sharing this!
Re: Step into in debugging.
On Wednesday, 22 February 2017 at 14:23:02 UTC, ANtlord wrote: My bad again. It appears I have a smaller knowledge that I think. I'm ended up in typecons because `allocator` is made by template `scoped` returns instance of internal static struct `Scoped` and it has alias is asigned to method `Scoped_payload` and this is called on https://github.com/ANtlord/DCD/blob/master/src/server/autocomplete.d#L341. SOLVED And one more. Method `Scoped_payload` returns instance of type passed as template argument for template method `scoped`. As long as method `generateAutocompleteTrees` requires second argument must be instance of `IAllocator`, object `allocator` "transforms" to instance of class implements `IAllocator` by alias of `Scoped_payload` assigned to `this`.
Re: Step into in debugging.
On Monday, 20 February 2017 at 13:55:57 UTC, ANtlord wrote:
Hello! I care for project DCD. I've wanted to fix some issue in
it very long time ago. But when I get a little bit of free time
I encounter one big problem for me.
I can't debug it using GDB. I can't go through call stack
because I get a number of issues. Long story short I've decided
to solve them one by one and I really need community's help.
First of them. I can't make `step into` staying on this code
line
https://github.com/ANtlord/DCD/blob/master/src/server/autocomplete.d#L341.
What I do:
1. I compile DCD project using dub with dmd `dub build
--compiler=dmd --build=debug --config=server --parallel`
2. I run `gdb `. I do it from project root
directory always `gdb dcd-server`.
3. Inside gdb. I do `break autocomplete.d:341`
4. Inside gdb. I do `run -p 6767` where `-p 6767` is DCD
parameter.
5. I launch dcd-client with parameters for calling desirable
code `dcd-client --port 6767 myfile.d -c 53`. File `myfile.d`
consists of `void main(string[] args) { int[] asd = [4,2];
asd[0]. }`. Yes, it has one line.
6. Inside gdb. I stay on desirable line and I do `step`.
7. Inside gdb. I am ended up in
`/usr/include/dlang/dmd/std/typecons.d:6097` instead function
`generateAutocompleteTrees` called on the line.
I attach this screenshot to be clear
https://www.dropbox.com/s/6i2hmbvqwdvopur/Screenshot_20170220_214526.png?dl=0
Thanks.
My bad again. It appears I have a smaller knowledge that I think.
I'm ended up in typecons because `allocator` is made by template
`scoped` returns instance of internal static struct `Scoped` and
it has alias is asigned to method `Scoped_payload` and this is
called on
https://github.com/ANtlord/DCD/blob/master/src/server/autocomplete.d#L341.
SOLVED
Re: Step into in debugging.
On Monday, 20 February 2017 at 13:55:57 UTC, ANtlord wrote:
Hello! I care for project DCD. I've wanted to fix some issue in
it very long time ago. But when I get a little bit of free time
I encounter one big problem for me.
I can't debug it using GDB. I can't go through call stack
because I get a number of issues. Long story short I've decided
to solve them one by one and I really need community's help.
First of them. I can't make `step into` staying on this code
line
https://github.com/ANtlord/DCD/blob/master/src/server/autocomplete.d#L341.
What I do:
1. I compile DCD project using dub with dmd `dub build
--compiler=dmd --build=debug --config=server --parallel`
2. I run `gdb `. I do it from project root
directory always `gdb dcd-server`.
3. Inside gdb. I do `break autocomplete.d:341`
4. Inside gdb. I do `run -p 6767` where `-p 6767` is DCD
parameter.
5. I launch dcd-client with parameters for calling desirable
code `dcd-client --port 6767 myfile.d -c 53`. File `myfile.d`
consists of `void main(string[] args) { int[] asd = [4,2];
asd[0]. }`. Yes, it has one line.
6. Inside gdb. I stay on desirable line and I do `step`.
7. Inside gdb. I am ended up in
`/usr/include/dlang/dmd/std/typecons.d:6097` instead function
`generateAutocompleteTrees` called on the line.
I attach this screenshot to be clear
https://www.dropbox.com/s/6i2hmbvqwdvopur/Screenshot_20170220_214526.png?dl=0
Thanks.
My bad. I haven't point version dmd. It is 2.072.0. But I get
same issue using 2.073.1, but instead
`/usr/include/dlang/dmd/std/typecons.d:6097` I'm ended up in
`/usr/include/dlang/dmd/std/typecons.d:6408`.
One more. Am I ended up in typecons because on of arguments of
function `generateAutocompleteTrees` is template?
Step into in debugging.
Hello! I care for project DCD. I've wanted to fix some issue in
it very long time ago. But when I get a little bit of free time I
encounter one big problem for me.
I can't debug it using GDB. I can't go through call stack because
I get a number of issues. Long story short I've decided to solve
them one by one and I really need community's help.
First of them. I can't make `step into` staying on this code line
https://github.com/ANtlord/DCD/blob/master/src/server/autocomplete.d#L341.
What I do:
1. I compile DCD project using dub with dmd `dub build
--compiler=dmd --build=debug --config=server --parallel`
2. I run `gdb `. I do it from project root
directory always `gdb dcd-server`.
3. Inside gdb. I do `break autocomplete.d:341`
4. Inside gdb. I do `run -p 6767` where `-p 6767` is DCD
parameter.
5. I launch dcd-client with parameters for calling desirable code
`dcd-client --port 6767 myfile.d -c 53`. File `myfile.d`
consists of `void main(string[] args) { int[] asd = [4,2];
asd[0]. }`. Yes, it has one line.
6. Inside gdb. I stay on desirable line and I do `step`.
7. Inside gdb. I am ended up in
`/usr/include/dlang/dmd/std/typecons.d:6097` instead function
`generateAutocompleteTrees` called on the line.
I attach this screenshot to be clear
https://www.dropbox.com/s/6i2hmbvqwdvopur/Screenshot_20170220_214526.png?dl=0
Thanks.
[Issue 11437] [Mago] No source for step-by-step debugging, only raw assembler.
https://issues.dlang.org/show_bug.cgi?id=11437 Rainer Schuetze r.sagita...@gmx.de changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #6 from Rainer Schuetze r.sagita...@gmx.de --- dmd fixed and workaround now released in Visual D 0.3.38 --
[Issue 11437] [Mago] No source for step-by-step debugging, only raw assembler.
https://d.puremagic.com/issues/show_bug.cgi?id=11437 --- Comment #5 from Rainer Schuetze r.sagita...@gmx.de 2014-01-03 03:16:59 PST --- While I have added a workaround in mago for the test case, it is actually an instance of issue 7634. -- Configure issuemail: https://d.puremagic.com/issues/userprefs.cgi?tab=email --- You are receiving this mail because: ---
[Issue 11437] [Mago] No source for step-by-step debugging, only raw assembler.
https://d.puremagic.com/issues/show_bug.cgi?id=11437 --- Comment #4 from Denis Shelomovskij verylonglogin@gmail.com 2014-01-02 17:31:28 MSK --- (In reply to comment #3) should be fixed by mago 0.9 in https://github.com/D-Programming-Language/visuald/releases/tag/v0.3.38beta1 Original testcase works now and everything was fine until some dmd changes. Now it is again no debug information available for most projects I tested, so I use an older dmd version. Testcase: http://deoma-cmd.ru/files/other/testcases/Issue-11437-2.7z -- Configure issuemail: https://d.puremagic.com/issues/userprefs.cgi?tab=email --- You are receiving this mail because: ---
[Issue 11437] [Mago] No source for step-by-step debugging, only raw assembler.
https://d.puremagic.com/issues/show_bug.cgi?id=11437 --- Comment #3 from Rainer Schuetze r.sagita...@gmx.de 2013-11-29 02:39:58 PST --- should be fixed by mago 0.9 in https://github.com/D-Programming-Language/visuald/releases/tag/v0.3.38beta1 -- Configure issuemail: https://d.puremagic.com/issues/userprefs.cgi?tab=email --- You are receiving this mail because: ---
[Issue 11437] [Mago] No source for step-by-step debugging, only raw assembler.
https://d.puremagic.com/issues/show_bug.cgi?id=11437 Rainer Schuetze r.sagita...@gmx.de changed: What|Removed |Added CC||r.sagita...@gmx.de --- Comment #2 from Rainer Schuetze r.sagita...@gmx.de 2013-11-06 00:40:07 PST --- I can reproduce the problem. It seems mago is as picky about the order of debug records as recommended by the CV4 documentation (but not made mandatory). As a workaround, the CVPACK utility is supposed to fix that, though I'm unsure if it can still be found somewhere. -- Configure issuemail: https://d.puremagic.com/issues/userprefs.cgi?tab=email --- You are receiving this mail because: ---
[Issue 11437] New: [Mago] No source for step-by-step debugging, only raw assembler.
http://d.puremagic.com/issues/show_bug.cgi?id=11437 Summary: [Mago] No source for step-by-step debugging, only raw assembler. Product: D Version: D2 Platform: All OS/Version: All Status: NEW Severity: major Priority: P2 Component: VisualD AssignedTo: nob...@puremagic.com ReportedBy: verylonglogin@gmail.com --- Comment #0 from Denis Shelomovskij verylonglogin@gmail.com 2013-11-04 17:54:07 MSK --- Sometimes Mago behaves like there is no debug information available. E.g. in attached testcase if lib_d.lib is removed from linker arguments or placed after GtkD.lib Mago works. Testcase: http://deoma-cmd.ru/files/other/testcases/Mago-error-1.7z -- Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email --- You are receiving this mail because: ---
[Issue 11437] [Mago] No source for step-by-step debugging, only raw assembler.
http://d.puremagic.com/issues/show_bug.cgi?id=11437 --- Comment #1 from Denis Shelomovskij verylonglogin@gmail.com 2013-11-04 18:02:48 MSK --- Workaround: Play with the order libraries are passed to the linker. I.e. change the order, press Step into (F11), repeat if you still see raw assembler. -- Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email --- You are receiving this mail because: ---
