cvs commit: src/sys/conf files.amd64 files.i386 src/sys/amd64/linux32 linux.h linux32_dummy.c linux32_machdep.c linux32_sysvec.c src/sys/compat/linux linux_misc.c linux_signal.c linux_util.c src/sys/i386/linux linux.h linux_dummy.c ...

[Bjoern A. Zeeb]

On Tue, 15 Aug 2006, Alexander Leidinger wrote:

> netchild    2006-08-15 12:54:30 UTC
>
>  FreeBSD src repository
>
>  Modified files:
>    sys/conf             files.amd64 files.i386
>    sys/amd64/linux32    linux.h linux32_dummy.c linux32_machdep.c
>                         linux32_sysvec.c
>    sys/compat/linux     linux_misc.c linux_signal.c linux_util.c
>    sys/i386/linux       linux.h linux_dummy.c linux_machdep.c
>                         linux_sysvec.c
>  Log:
>  Add the linux 2.6.x stuff (not used by default!):
>   - TLS - complete
>   - pid/tid mangling - complete
>   - thread area - complete
>   - futexes - complete with issues
>   - clone() extension - complete with some possible minor issues
>   - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is

...

>  On amd64 not everything is supported like on i386, the catchup is planned for
>  later when the remaining bugs in the new functions are fixed.

Yeah I like that.  I am doing some of that catchup almost 8 years later and find that the supposedly really simple clock_* wrappers in i386 are completely wrong and unusable  and the chances are that this has contributed to quite a few unnoticed problems over time, given we only get weird runtime errors rather than an "unimplemented syscall" every time.

Just saying as my linux compat code frustration levels are rising again.  *sigh*


-- 
Bjoern A. Zeeb             "Come on. Learn, goddamn it.", WarGames, 1983