> Date: Thu, 03 Dec 2009 00:30:10 -0800
> From: Mike Shapiro <mws at sun.com>
> Subject: Re: delete obsolete system call traps [PSARC/2009/657]
> To: "Roger A. Faulkner" <Roger.Faulkner at sun.com>
> Cc: psarc-ext at sac.sfbay.sun.com, Bart.Smaalders at sun.com,
Bryan.Cantrill at sun.com, Michael.Shapiro at sun.com, Darrin.Johnson at sun.com
>
...
> I certainly agree some of this is implementation and doesn't matter.
> And from a technical perspective, because the current DTrace syscall
> provider advertises the sysent table names and not the sub-coded man
> page (2) view of system calls (something we intend to rectify
> someday), we marked those names as Private in our stability mechanism.
>
> But from a practical perspective, I'm very uncomfortable with the *at
> part of this project, whereby extremely well-known fundamental names
> used extensively by users of DTrace based on long-standing
> fundamentals like syscall::open:entry just go away, essentially for no
> benefit.
I discussed this matter offline with Michael Shapiro and Bryan Cantrill.
While it is unfortunate that the current DTrace syscall provider
has caused the syscall trap table to have become a public interface,
DTrace usage is widespread enough to warrent keeping unchanged what
Mike refers to as "extremely well-known fundamental names used
extensively by users of DTrace based on long-standing fundamentals
like syscall::open:entry".
This not a tenable long-term solution. The kernel <-> user-level
syscall trap interface has never been a Solaris public interface.
The long-term solution is to implement what Mike refers to as
"the sub-coded man page (2) view of system calls" in the DTrace
syscall provider.
In the interim, for the current PSARC case for deleting obsolete
system call traps, I am changing the proposal as follows:
===========================================================
Retain these old system call traps;
make sure they are called from libc as always so that
existing DTrace scripts continue to work as always:
open()
open64()
stat()
lstat()
fstat()
stat64()
lstat64()
fstat64()
rename()
unlink()
rmdir()
chown()
lchown()
fchown()
access()
Delete these in favor of open*(path, O_WRONLY | O_CREAT | O_TRUNC, mode):
creat()
creat64()
Delete this in favor of fcntl(F_DUPFD):
dup()
Delete these from i386 code in favor of *stat() amd mknod()
(they should have been deleted when COFF support was dropped):
_xstat()
_lxstat()
_fxstat()
_xmknod()
Delete these (they are never called from libc):
exec()
fork1()
forkall()
poll()
utime()
utimes()
umount()
wait()
lwp_mutex_lock()
lwp_sema_wait()
Delete this:
fsat()
in favor of individual (new) system call traps for each of its subcodes:
openat()
openat64()
fstatat()
fstatat64()
renameat()
unlinkat()
fchownat()
faccessat()
===========================================================
Roger Faulkner
