I am getting this weird linker error when compiling my code; what does it mean?

Jeroen Bollen:

> I am getting this weird linker error when compiling my code; what does it mean?

If your code used to work, and you have just tried dmd 2.064 then as try to compile with -allinst.

Bye,
bearophile

On Saturday, 23 November 2013 at 21:32:13 UTC, bearophile wrote:
> Jeroen Bollen:
>
>> I am getting this weird linker error when compiling my code; what does it mean?
>
> If your code used to work, and you have just tried dmd 2.064 then as try to compile with -allinst.
>
> Bye,
> bearophile

No luck with that flag.

Jeroen Bollen:

> I am getting this weird linker error when compiling my code; what does it mean?

Is your code not compiling since switching to a newer compiler? What system and compilers are you using? How many modules are in your program, and what build strategy/software are you using?

Bye,
bearophile

On Saturday, 23 November 2013 at 21:05:59 UTC, Jeroen Bollen wrote:
> I am getting this weird linker error when compiling my code; what does it mean?

I saw this on stackoverflow first and answered there, let me link it:

http://stackoverflow.com/questions/20168136/linker-error-cannot-link-d16typeinfo-hayayaa6-initz/20168874#20168874

I've been seeing this problem on and off for years now, but can never get it minimized to a good test case. I think the order of files on the compile command line matter along with other hard to track down factors. So I just hack around it and move on.

Not sure if the type in there is right, I demangled it by eyeball (why wouldn't the core.demangle work on that? it isn't even that complex of a type, if I'm right...)

On Saturday, 23 November 2013 at 22:34:54 UTC, bearophile wrote:
> Jeroen Bollen:
>
>> I am getting this weird linker error when compiling my code; what does it mean?
>
> Is your code not compiling since switching to a newer compiler? What system and compilers are you using? How many modules are in your program, and what build strategy/software are you using?
>
> Bye,
> bearophile

It's the first time I compiled it. I am using a regular compile command

> dmd -m64 -of"../Build/main" (lots of files here)

Linux 64 bits

On Saturday, 23 November 2013 at 22:49:54 UTC, Adam D. Ruppe wrote:
> On Saturday, 23 November 2013 at 21:05:59 UTC, Jeroen Bollen wrote:
>> I am getting this weird linker error when compiling my code; what does it mean?
>
> I saw this on stackoverflow first and answered there, let me link it:
>
> http://stackoverflow.com/questions/20168136/linker-error-cannot-link-d16typeinfo-hayayaa6-initz/20168874#20168874
>
> I've been seeing this problem on and off for years now, but can never get it minimized to a good test case. I think the order of files on the compile command line matter along with other hard to track down factors. So I just hack around it and move on.
>
> Not sure if the type in there is right, I demangled it by eyeball (why wouldn't the core.demangle work on that? it isn't even that complex of a type, if I'm right...)

Flagged as Answer

I added the code to my GitHub repo; there don't seem to be any uncommon associative arrays: https://github.com/SanePumpkins/FastCGI.D

On Saturday, 23 November 2013 at 23:30:09 UTC, Jeroen Bollen wrote:
> I added the code to my GitHub repo; there don't seem to be any uncommon associative arrays:

Yea, it is the immutable string[string], I used the same pattern in my cgi.d and saw that too (sometimes, like I said, it is randomly there or not. it seems to be the outer immutable that triggers the problem.

BTW, my cgi.d has fastcgi support too
https://github.com/adamdruppe/misc-stuff-including-D-programming-language-web-stuff/blob/master/cgi.d

compile with -version=fastcgi. It uses the C libfcgi in that mode. Feel free to copy any of my code you want if it is helpful to you.

On Saturday, 23 November 2013 at 23:47:11 UTC, Adam D. Ruppe wrote:
> On Saturday, 23 November 2013 at 23:30:09 UTC, Jeroen Bollen wrote:
>> I added the code to my GitHub repo; there don't seem to be any uncommon associative arrays:
>
> Yea, it is the immutable string[string], I used the same pattern in my cgi.d and saw that too (sometimes, like I said, it is randomly there or not. it seems to be the outer immutable that triggers the problem.
>
> BTW, my cgi.d has fastcgi support too
> https://github.com/adamdruppe/misc-stuff-including-D-programming-language-web-stuff/blob/master/cgi.d
>
> compile with -version=fastcgi. It uses the C libfcgi in that mode. Feel free to copy any of my code you want if it is helpful to you.

Your work actually inspired mine. I saw it used some external code so I am basically writing it without those. Somewhere in progress of that I managed to completely throw over the structure of the code, and well yeah, now I ended up with that. :P