Re: Something wrong with GC
On 22.03.2016 16:56, ag0aep6g wrote: I've filed an issue: https://issues.dlang.org/show_bug.cgi?id=15821 And it's been fixed: https://github.com/D-Programming-Language/druntime/pull/1519 Since the issue was a regression, the fix was made against the stable branch. It's going to be in the next release. But it's not yet in master, which means it's also not going to be in the nightlies for now.
Re: Something wrong with GC
On Tuesday, 22 March 2016 at 13:46:41 UTC, stunaep wrote: So what am I do to? Just learn more about available containers and their semantics. Maybe you don't need Array!T when there is a simple T[]. If you think you do need Array, then think about memory management: where are you going to allocate the data - in the GC heap or outside it. Depending on your answers there are different approaches. It's all solvable if you pause and think what exactly you're trying to do.
Re: Something wrong with GC
On 20.03.2016 08:49, stunaep wrote:
The gc throws invalid memory errors if I use Arrays from std.container.
For example, this throws an InvalidMemoryOperationError:
import std.stdio;
import std.container;
void main() {
new Test();
}
class Test {
private Array!string test = Array!string();
this() {
test.insert("test");
writeln(test[0]);
}
}
I can reproduce the InvalidMemoryOperationError with git head dmd, but
there doesn't seem to be a problem with 2.070. So I'd say this is a
regression in the development version.
I've filed an issue: https://issues.dlang.org/show_bug.cgi?id=15821
You're probably building dmd/phobos from git, or you're using a nightly,
right? Maybe you can go back to 2.070.2 until this is sorted out.
Re: Something wrong with GC
On Tuesday, 22 March 2016 at 13:46:41 UTC, stunaep wrote:
public class Example2 {
private int one;
private int two;
public this(int one, int two) {
this.one = one;
this.two = two;
}
}
in a tree map and list of some sort. Neither of the above work
whether they are classes or structs and it's starting to become
quite bothersome...
Is there a particular reason why you don't want to use the
standard ranges?
public class Example2 {
private int one;
private int two;
public this(int one, int two) {
this.one = one;
this.two = two;
}
}
void main()
{
auto myExamplesList = [ new Example2( 6,3 ), new Example2(7,5) ];
// Note that if you do a lot of appending then using
Appender is more performant than ~=
myExamplesList ~= new Example2(9,1);
}
For trees there is also redBlackTree
Re: Something wrong with GC
On Monday, 21 March 2016 at 07:55:39 UTC, thedeemon wrote:
On Sunday, 20 March 2016 at 07:49:17 UTC, stunaep wrote:
The gc throws invalid memory errors if I use Arrays from
std.container.
Those arrays are for RAII-style deterministic memory release,
they shouldn't be freely mixed with GC-allocated things. What
happens here is while initializing Array sees it got some GC-ed
value type (strings), so it tells GC to look after those
strings. When your program ends runtime does a GC cycle, finds
your Test object, calls its destructor that calls Array
destructor that tries to tell GC not to look at its data
anymore. But during a GC cycle it's currently illegal to call
such GC methods, so it throws an error.
Moral of this story: try not to store "managed" (collected by
GC) types in Array and/or try not to have Arrays inside
"managed" objects. If Test was a struct instead of a class, it
would work fine.
So what am I do to? Any other language can do such a thing so
trivially... I also run into the same problem with
emsi_containers TreeMap. It is imperative that I can store data
such as
public class Example1 {
private File file;
public this(File f) {
this.file = f;
}
}
or
public class Example2 {
private int one;
private int two;
public this(int one, int two) {
this.one = one;
this.two = two;
}
}
in a tree map and list of some sort. Neither of the above work
whether they are classes or structs and it's starting to become
quite bothersome...
Re: Something wrong with GC
On Sunday, 20 March 2016 at 07:49:17 UTC, stunaep wrote: The gc throws invalid memory errors if I use Arrays from std.container. ... Not sure what to do here I don't know what your larger goal is, but maybe std.array.Appender would be a better fit?
Re: Something wrong with GC
On Sunday, 20 March 2016 at 07:49:17 UTC, stunaep wrote: The gc throws invalid memory errors if I use Arrays from std.container. Those arrays are for RAII-style deterministic memory release, they shouldn't be freely mixed with GC-allocated things. What happens here is while initializing Array sees it got some GC-ed value type (strings), so it tells GC to look after those strings. When your program ends runtime does a GC cycle, finds your Test object, calls its destructor that calls Array destructor that tries to tell GC not to look at its data anymore. But during a GC cycle it's currently illegal to call such GC methods, so it throws an error. Moral of this story: try not to store "managed" (collected by GC) types in Array and/or try not to have Arrays inside "managed" objects. If Test was a struct instead of a class, it would work fine.
