ok - according to unistd.h we now have exactly 256 syscalls allocated (unless im missing something). my code needs to be able to account for every single possible syscall and so i need to be able to store the syscall number in a standard way. not all syscalls are catered for on the outset by at any time the user can say "i need to use syscall x which takes y parameters" and the code will be able to take care of it.
the problem is that i am currently reserving only 8 bits for the syscall number. this is ok for now but if we ever get another syscall its going to be unuseable by my existing code :) - should i be reserving 16 bits now in preperation for some new syscalls being added ?
On Tue, 2002-04-09 at 04:10, mark manning wrote:
> ok - according to unistd.h we now have exactly 256 syscalls allocated (unless im missing something). my code needs to be able to account for every single possible syscall and so i need to be able to store the syscall number in a standard way. not all syscalls are catered for on the outset by at any time the user can say "i need to use syscall x which takes y parameters" and the code will be able to take care of it.
>
> the problem is that i am currently reserving only 8 bits for the syscall number. this is ok for now but if we ever get another syscall its going to be unuseable by my existing code :) - should i be reserving 16 bits now in preperation for some new syscalls being added ?
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
Hmm...
dunno if you got this right. There are maximal 256 syscalls possible,
and, right -- exactly this amount of syscalls is in the entrytable. But
alotalotalot of them are defined as sys_ni_syscall (not yet
implemented).
I think there is still some space for enhancements. See
arch/i386/kernel/entry.S.
Regards
Frank
Hi all!
I've got a really big problem with kernel 2.4.18 and 2.4.17 on
an Alpha.
I can compile, install and boot the kernel on my Alpha.
But if I shutdown the machine without shutting down the
system - I know this is crazy, but sometimes this happens...
So if I'm this mad and restart the machine afterwards, I get
a lot of fsck errors 'til the system give up and tell's me,
that I have to check it with fsck myself.
OK, I did so... fsck -y /dev/sda1 -> Works perfectly. After
fsck has corrected more than 1000 errors I'm able to
reboot the machine.
And than: MY SYSTEM IS NO LONGER BOOTABLE. It's totally
currupted...
I never saw the filesystem curruption bug on Intel, but it
sounds like this.
Is this the same bug that was on Intel?
Have I done something wrong?
Are there any alpha-users in this list? :o))))
Greetz, I look forward for answers,
Oliver
>-----Original Message-----
>From: [email protected]
>[mailto:[email protected]]On Behalf Of Oliver Pitzeier
>Sent: Tuesday, April 09, 2002 12:20 PM
>To: [email protected]
>Subject: Compaq Alpha DS10 - Kernel 2.4.18
>
>
>Hi all!
>
>I've got a really big problem with kernel 2.4.18 and 2.4.17 on
>an Alpha.
>
>I can compile, install and boot the kernel on my Alpha.
>But if I shutdown the machine without shutting down the
>system - I know this is crazy, but sometimes this happens...
>
>So if I'm this mad and restart the machine afterwards, I get
>a lot of fsck errors 'til the system give up and tell's me,
>that I have to check it with fsck myself.
>
>OK, I did so... fsck -y /dev/sda1 -> Works perfectly. After
>fsck has corrected more than 1000 errors I'm able to
>reboot the machine.
>
>And than: MY SYSTEM IS NO LONGER BOOTABLE. It's totally
>currupted...
>
>I never saw the filesystem curruption bug on Intel, but it
>sounds like this.
>
>Is this the same bug that was on Intel?
>Have I done something wrong?
>
>Are there any alpha-users in this list? :o))))
>
>Greetz, I look forward for answers,
> Oliver
>
>
>-
Think about one of journaling filesystems: ext3, reiserfs, xfs, jfs...
Regards,
Damian Wrobel
On Tuesday 09 April 2002 11:19 am, Oliver Pitzeier wrote:
> Hi all!
> [...]
> OK, I did so... fsck -y /dev/sda1 -> Works perfectly. After
> fsck has corrected more than 1000 errors I'm able to
> reboot the machine.
>
> And than: MY SYSTEM IS NO LONGER BOOTABLE. It's totally
> currupted...
> Are there any alpha-users in this list? :o))))
$uname -a
Linux frumious.unidec.co.uk 2.4.19pre1 #2 Wed Feb 27 14:45:16 GMT 2002 alpha
unknown
Yup. Not had any problems despite a series of recent powercuts (including
one where the mains voltage dropped from ~230V to ~150V), but did have very
long fsck()s afterwards. So I gave up and went to ext3 - all it takes is
kernel support and a quick tune2fs -j /dev/XXX ;-)
cheers
john
On 9 Apr 2002, Frank Schaefer wrote:
| On Tue, 2002-04-09 at 04:10, mark manning wrote:
| > ok - according to unistd.h we now have exactly 256 syscalls allocated (unless im missing something). my code needs to be able to account for every single possible syscall and so i need to be able to store the syscall number in a standard way. not all syscalls are catered for on the outset by at any time the user can say "i need to use syscall x which takes y parameters" and the code will be able to take care of it.
| >
| > the problem is that i am currently reserving only 8 bits for the syscall number. this is ok for now but if we ever get another syscall its going to be unuseable by my existing code :) - should i be reserving 16 bits now in preperation for some new syscalls being added ?
| > -
| >
| Hmm...
|
| dunno if you got this right. There are maximal 256 syscalls possible,
| and, right -- exactly this amount of syscalls is in the entrytable. But
| alotalotalot of them are defined as sys_ni_syscall (not yet
| implemented).
| I think there is still some space for enhancements. See
| arch/i386/kernel/entry.S.
Where is the limitation of 256 syscalls possible?
--
~Randy