Messages in this thread | | | Subject | Re: Will IPC changes be included in the mainstream kernel ? | Date | Fri, 18 Dec 1998 15:41:36 +0200 (EET) | From | Matti Aarnio <> |
| |
Serguei Koubouchine <ksi@gu.net> is asking: > > Linus, can you tell us, mere mortals, will the IPC changes (new msg.h etc.) > be included in the mainstream kernel? I have to make a decision what to > patch, ac kernels or all the programs affected by those changes....
(I am not Linus, but..)
Sergei, Do you mean that you are compiling code in pre-GLIBC-2.0 environment which uses <linux/msg.h> in user programs ?
In GLIBC 2.0/2.1 the <linux/msg.h> header should *not* be included by user-space software, and thus you should not have any grievances about this issue. (See <sys/msg.h> )
The new interface even is binary compatible with old *binaries*.
What happens to the 'u_short' fields when the amount of data in IPC queue exceeds 'u_short' value range is that apparently the real value is stored there modulo 'u_short' value range.. (e.g. 128k+10 bytes will appear to your old binaries as 10 bytes. Not as - for example - USHORT_MAX value.) Yeah, suboptimal, somewhat dangerous even.
> If the changes are included, the whole bunch of programs are broken. It's > fixable, but I don't want to waste my time on an unnecessary work. If the > changes are NOT included, the acXX patches has to be reworked to get the old > IPC back...
Where did it go ? I see it still.
> Can you inform us of your decision ? > ======================================================================= > Serguei Koubouchine aka the Tamer < > The impossible we do immediately. > e-mail: ksi@gu.net SK320-RIPE < > Miracles require 24-hour notice.
/Matti Aarnio <matti.aarnio@sonera.fi>
Anybody interested in files over 2G in size in i386 systems ? Contact me, dangerous work-in-progress code soon available for brave experimenters for 2.1.131+ at: ftp://mea.tmt.tele.fi/linux/lfs-patches-21131ac-v**.diff (old stat() yields ULONG_MAX for filesizes exceeding 4G-1 at 32-bit systems..)
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |