Thank you for your help! Yes with your pointer I found out that cc1 writes a .s file in the /tmp directory, and then the assembler /usr/bin/as opens that for reading using fopen() and outputs foobar.o (without any renaming). Still I don't know what the assembler is using to open foobar.o for writing (strace says open() but I intercept that and any variants and none of them are used, maybe the assembler is built with syscalls linked statically). So I guess I need to understand the assembler source and/or how it links the system calls. I can see that the assembler is not part of gcc so I guess that is where this thread ends. Thank you for your help and if you have anything more to teach me, please do, thank you ----- Original Message ----- From: Manuel López-Ibáñez <lopezibanez@xxxxxxxxx> To: Mark Galeck <mark_galeck@xxxxxxxxxxx>; "gcc-help@xxxxxxxxxxx" <gcc-help@xxxxxxxxxxx> Sent: Saturday, March 5, 2016 3:52 AM Subject: Re: where in gcc sources is foobar.o written? On 05/03/16 11:23, Mark Galeck wrote: > Hello, > > > as you know, the source of gcc is quite complex, especially for someone who managed to go through their MS in Computer Science without taking a compiler class :) > > Where in the sources is the file foobar.o opened for writing in The foobar.o in your command-line is created by the assembler. Append '-###' to see all programs invoked by the gcc driver. See https://gcc.gnu.org/wiki/DebuggingGCC It is possible that the assembler creates the .o file in some /tmp/ directory, then renames/moves it to the target directory. There is a lot of useful info in https://gcc.gnu.org/wiki/GettingStarted Unfortunately, it is badly organized. Any help in organizing it would be appreciated. Cheers, Manuel.
