On 02-Feb-01 Peter Pentchev wrote:
> On Thu, Feb 01, 2001 at 01:19:24PM -0800, John Baldwin wrote:
>>
>> On 01-Feb-01 Doug White wrote:
>> > On Mon, 29 Jan 2001, John Baldwin wrote:
>> >
>> >> Then only rename it in 4.x We can do an API change for 5.0. We'll be
>> >> renaming syscall2() back to syscall() in 5.0 for example. We don't
>> >> want to end up with syscall47() someday in FreeBSD 67.2. :-P
>> >
>> > And what happens to apps using the previous syscall(2) interface? They
>> > die horribly? That's not acceptable.
>>
>> Huh? syscall2() is an internal kernel function. The only thing that might
>> call it is a kernel module, and we won't support 4.x KLD's on 5.0. Period.
>> It was renamed to syscall2() to cause old KLD's to fail to load with a
>> symbol
>> lookup problem back when teh MPSAFE flag was first added to x86 syscalls.
>> The
>> only thing that changing it back to syscall() in -current does is cause any
>> 6
>> month old 5.0-current KLD's that happen to call syscall() to call it wrong.
>
> Uhm.. I believe he meant syscall(2) as in 'the "syscall" entity described
> in section 2 of the manual', which in this case happens to be:
>
> NAME
> syscall, __syscall - indirect system call
>
> ..as used in src/lib/libc/sys/*.c.
I'm not renaming that function. I'm renaming syscall2() in
sys/i386/i386/trap.c, which is an internal function only used
by the kernel for the kernel side of syscall entry.
--
John Baldwin <[EMAIL PROTECTED]> -- http://www.FreeBSD.org/~jhb/
PGP Key: http://www.baldwin.cx/~john/pgpkey.asc
"Power Users Use the Power to Serve!" - http://www.FreeBSD.org/
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
