Hi...
> > Hi, I did a quick look at the code of the NPTL. I found that the
> > pthread_self does not return the tid but the address of the struct
> > pthread which is allocated in the stack (at least on i386 and
> > possibly on other Arch too). Struct pthread contains various
> > information used by the pthreaded calls. It created in the stack
> > during pthread creation. pthread_self just finds the structure from
> > stack and return its address.
Great, thanks for sharing the info :)
> clone(child_stack=0xb7f7d4d4,
> flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|
> CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|
> CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID,
> parent_tidptr=0xb7f7dbe8,
> {entry_number:6, base_addr:0xb7f7dba0, limit:1048575,
> seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1,
> seg_not_present:0, useable:1},
> child_tidptr=0xb7f7dbe8) = 21945
hm, IIRC, what you see there (base_addr) is base address of TLS segment.
This segment is mapped on certain entry (see entry_number) in local CPU
GDT. Also, this segment is separated for user mode stack. However, it
makes sense to use base_addr of TLS segment since I must unique for
each thread on the same thread group.
regards
Mulyadi
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
FAQ: http://kernelnewbies.org/faq/
