* The world's worst RPM
@ 1998-01-14 6:43 Alex deVries
1998-01-14 10:28 ` ralf
1998-01-14 13:28 ` Oliver Frommel
0 siblings, 2 replies; 22+ messages in thread
From: Alex deVries @ 1998-01-14 6:43 UTC (permalink / raw)
To: SGI Linux
I have generated the world's worst RPM for util-linux, which is much
needed, since it contains fdisk.
I had to kludge some things; util-linux is truly an awful thing. It
contains things like hwclock, which sets the clock on ISA bus only
machines. This is designed only for sparc, alpha and i386 _ONLY_.
Anyway, I kludged it into compiling, but I really had to guess on the
parameters for the disk operations. Fdisk compiles and works, but I can't
really test it as I need both disks on my system. Could somebody give it
a try?
Also, it has things like more , write and mkswap. Yay! I now have swap
space!
So, give it a try, or clean up my diff files.
- Alex
--
Alex deVries Run Linux on everything,
System Administrator run everything on Linux.
The EngSoc Project Send spam to spam@engsoc.carleton.ca.
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: The world's worst RPM 1998-01-14 6:43 The world's worst RPM Alex deVries @ 1998-01-14 10:28 ` ralf 1998-01-14 16:38 ` Alex deVries 1998-01-14 13:28 ` Oliver Frommel 1 sibling, 1 reply; 22+ messages in thread From: ralf @ 1998-01-14 10:28 UTC (permalink / raw) To: Alex deVries; +Cc: SGI Linux On Wed, Jan 14, 1998 at 01:43:00AM -0500, Alex deVries wrote: > I have generated the world's worst RPM for util-linux, which is much > needed, since it contains fdisk. > > I had to kludge some things; util-linux is truly an awful thing. It > contains things like hwclock, which sets the clock on ISA bus only > machines. This is designed only for sparc, alpha and i386 _ONLY_. You must be talking about clock. As I remember hwclock is using some /dev/rtc and should be portable. Anyway, all the world is using the same rtc from Dallas Semiconductor, so ... > Anyway, I kludged it into compiling, but I really had to guess on the > parameters for the disk operations. Fdisk compiles and works, but I can't > really test it as I need both disks on my system. Could somebody give it > a try? > > Also, it has things like more , write and mkswap. Yay! I now have swap > space! > > So, give it a try, or clean up my diff files. Suicide doesn't look like a bad alternative to cleaning the diffs. And it's not your fault. Actually I rememer that somebody already put his mental health on the game, there is a glibc patch for util-linux floating around. Ralf ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 10:28 ` ralf @ 1998-01-14 16:38 ` Alex deVries 0 siblings, 0 replies; 22+ messages in thread From: Alex deVries @ 1998-01-14 16:38 UTC (permalink / raw) To: ralf; +Cc: SGI Linux On Wed, 14 Jan 1998 ralf@uni-koblenz.de wrote: > On Wed, Jan 14, 1998 at 01:43:00AM -0500, Alex deVries wrote: > > I have generated the world's worst RPM for util-linux, which is much > > needed, since it contains fdisk. > > I had to kludge some things; util-linux is truly an awful thing. It > > contains things like hwclock, which sets the clock on ISA bus only > > machines. This is designed only for sparc, alpha and i386 _ONLY_. > You must be talking about clock. As I remember hwclock is using some > /dev/rtc and should be portable. Anyway, all the world is using the > same rtc from Dallas Semiconductor, so ... On closer inspection it's just a matter of doing something like: #ifdef ISA_BUS (ISA specific routines that use iopl()) #endif Yes, it is fixable. > Suicide doesn't look like a bad alternative to cleaning the diffs. And > it's not your fault. Actually I rememer that somebody already put his > mental health on the game, there is a glibc patch for util-linux floating > around. Yeah. Those are in the source RPM, and I'm using them. There's a lot of #ifdef __SPARC__ or similiar statements that need to be fixed for mips. I really think this should be broken into multiple packages. - Alex ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 6:43 The world's worst RPM Alex deVries 1998-01-14 10:28 ` ralf @ 1998-01-14 13:28 ` Oliver Frommel 1998-01-14 13:58 ` ralf 1998-01-14 14:36 ` Alan Cox 1 sibling, 2 replies; 22+ messages in thread From: Oliver Frommel @ 1998-01-14 13:28 UTC (permalink / raw) To: Alex deVries; +Cc: SGI Linux > > Anyway, I kludged it into compiling, but I really had to guess on the > parameters for the disk operations. Fdisk compiles and works, but I can't > really test it as I need both disks on my system. Could somebody give it > a try? > i'll do that tomorrow (when i am back at my indy). i also wonder what you changed to not only make it compile but also work, but i guess i'll see .. :) anyway if there's still anything to work on, converning fdisk, i'd have some time to do so this weekend .. oliver ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 13:28 ` Oliver Frommel @ 1998-01-14 13:58 ` ralf 1998-01-14 16:42 ` Alex deVries 1998-01-14 14:36 ` Alan Cox 1 sibling, 1 reply; 22+ messages in thread From: ralf @ 1998-01-14 13:58 UTC (permalink / raw) To: Oliver Frommel; +Cc: Alex deVries, SGI Linux On Wed, Jan 14, 1998 at 02:28:32PM +0100, Oliver Frommel wrote: > > Anyway, I kludged it into compiling, but I really had to guess on the > > parameters for the disk operations. Fdisk compiles and works, but I can't > > really test it as I need both disks on my system. Could somebody give it > > a try? > > > > i'll do that tomorrow (when i am back at my indy). > i also wonder what you changed to not only make it compile but also work, > but i guess i'll see .. :) > anyway if there's still anything to work on, converning fdisk, i'd have some > time to do so this weekend .. As I understand Alex's words he only made fdisk running but it still doesn't support IRIX style disklabels. How about that? That's aside of getting rid of sash one of the major thing we still need to do in order to get fully self hosting. Ralf ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 13:58 ` ralf @ 1998-01-14 16:42 ` Alex deVries 0 siblings, 0 replies; 22+ messages in thread From: Alex deVries @ 1998-01-14 16:42 UTC (permalink / raw) To: ralf; +Cc: Oliver Frommel, SGI Linux On Wed, 14 Jan 1998 ralf@uni-koblenz.de wrote: > As I understand Alex's words he only made fdisk running but it still doesn't > support IRIX style disklabels. How about that? That's correct. It should be able to modify partition tables of disks, but Irix won't recognize it. That's no problem if you have no intention of using Irix once your machine is running smoothly, like me. - Alex ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 13:28 ` Oliver Frommel 1998-01-14 13:58 ` ralf @ 1998-01-14 14:36 ` Alan Cox 1998-01-14 14:29 ` ralf 1998-01-14 14:36 ` Alan Cox 1 sibling, 2 replies; 22+ messages in thread From: Alan Cox @ 1998-01-14 14:36 UTC (permalink / raw) To: Oliver Frommel; +Cc: adevries, linux I've stuck some more new rpms on ftp.uk.linux.org incoming and SGI dir including sox, xmorph, ncompress, sysklogd Alan ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 14:36 ` Alan Cox @ 1998-01-14 14:29 ` ralf 1998-01-14 15:07 ` Alan Cox 1998-01-14 14:36 ` Alan Cox 1 sibling, 1 reply; 22+ messages in thread From: ralf @ 1998-01-14 14:29 UTC (permalink / raw) To: Alan Cox; +Cc: Oliver Frommel, adevries, linux On Wed, Jan 14, 1998 at 02:36:17PM +0000, Alan Cox wrote: > I've stuck some more new rpms on ftp.uk.linux.org incoming and SGI dir > including sox, xmorph, ncompress, sysklogd Have you taken care of the fact that ncompress makes assumptions about the byteorder of the machine it's running on? For MIPS it's broken, but I don't remember if it was on big or little endian ... Ralf ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 14:29 ` ralf @ 1998-01-14 15:07 ` Alan Cox 1998-01-14 20:58 ` Thomas Bogendoerfer 0 siblings, 1 reply; 22+ messages in thread From: Alan Cox @ 1998-01-14 15:07 UTC (permalink / raw) To: ralf; +Cc: alan, oliver, adevries, linux > Have you taken care of the fact that ncompress makes assumptions about > the byteorder of the machine it's running on? For MIPS it's broken, but > I don't remember if it was on big or little endian ... Its passed in the spec file. I've even port mipsel in as well for you ;) ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 15:07 ` Alan Cox @ 1998-01-14 20:58 ` Thomas Bogendoerfer 1998-01-14 22:17 ` Alan Cox 1998-01-14 23:59 ` The world's worst RPM Miguel de Icaza 0 siblings, 2 replies; 22+ messages in thread From: Thomas Bogendoerfer @ 1998-01-14 20:58 UTC (permalink / raw) To: linux On Wed, Jan 14, 1998 at 03:07:27PM +0000, Alan Cox wrote: > > Have you taken care of the fact that ncompress makes assumptions about > > the byteorder of the machine it's running on? For MIPS it's broken, but > > I don't remember if it was on big or little endian ... > > Its passed in the spec file. I've even port mipsel in as well for you ;) hmm, to do this with only one src.rpm, we need a little support from rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, that for .spec execution mips is defined for bot mipsel and mipseb, because there are changes, which work for both and we only need to seperate changes like that needed by ncompress. Comments ? Does anybody how to do this ? Thomas. -- See, you not only have to be a good coder to create a system like Linux, you have to be a sneaky bastard too ;-) [Linus Torvalds in <4rikft$7g5@linux.cs.Helsinki.FI>] ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 20:58 ` Thomas Bogendoerfer @ 1998-01-14 22:17 ` Alan Cox 1998-01-14 22:17 ` Alan Cox 1998-01-14 22:25 ` Alex deVries 1998-01-14 23:59 ` The world's worst RPM Miguel de Icaza 1 sibling, 2 replies; 22+ messages in thread From: Alan Cox @ 1998-01-14 22:17 UTC (permalink / raw) To: Thomas Bogendoerfer; +Cc: linux > hmm, to do this with only one src.rpm, we need a little support from > rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, > that for .spec execution mips is defined for bot mipsel and mipseb, because > there are changes, which work for both and we only need to seperate changes > like that needed by ncompress. Comments ? Does anybody how to do this ? Just ask Erik Troan nicely ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 22:17 ` Alan Cox @ 1998-01-14 22:17 ` Alan Cox 1998-01-14 22:25 ` Alex deVries 1 sibling, 0 replies; 22+ messages in thread From: Alan Cox @ 1998-01-14 22:17 UTC (permalink / raw) To: Thomas Bogendoerfer; +Cc: linux > hmm, to do this with only one src.rpm, we need a little support from > rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, > that for .spec execution mips is defined for bot mipsel and mipseb, because > there are changes, which work for both and we only need to seperate changes > like that needed by ncompress. Comments ? Does anybody how to do this ? Just ask Erik Troan nicely ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 22:17 ` Alan Cox 1998-01-14 22:17 ` Alan Cox @ 1998-01-14 22:25 ` Alex deVries 1998-01-15 0:03 ` ralf 1 sibling, 1 reply; 22+ messages in thread From: Alex deVries @ 1998-01-14 22:25 UTC (permalink / raw) To: Alan Cox; +Cc: Thomas Bogendoerfer, linux On Wed, 14 Jan 1998, Alan Cox wrote: > > hmm, to do this with only one src.rpm, we need a little support from > > rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, > > that for .spec execution mips is defined for bot mipsel and mipseb, because > > there are changes, which work for both and we only need to seperate changes > > like that needed by ncompress. Comments ? Does anybody how to do this ? > Just ask Erik Troan nicely He has already agreed to it, I just need to submit my patches to RPM that do the detection of the byte order for setting the soft-coded default architecture. It'll be done by the end of the week. - Alex ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 22:25 ` Alex deVries @ 1998-01-15 0:03 ` ralf 1998-01-15 3:56 ` mipseb and mipsel patches to RPM Alex deVries 0 siblings, 1 reply; 22+ messages in thread From: ralf @ 1998-01-15 0:03 UTC (permalink / raw) To: Alex deVries; +Cc: Alan Cox, Thomas Bogendoerfer, linux On Wed, Jan 14, 1998 at 05:25:31PM -0500, Alex deVries wrote: > On Wed, 14 Jan 1998, Alan Cox wrote: > > > hmm, to do this with only one src.rpm, we need a little support from > > > rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, > > > that for .spec execution mips is defined for bot mipsel and mipseb, because > > > there are changes, which work for both and we only need to seperate changes > > > like that needed by ncompress. Comments ? Does anybody how to do this ? > > Just ask Erik Troan nicely > > He has already agreed to it, I just need to submit my patches to RPM that > do the detection of the byte order for setting the soft-coded default > architecture. It'll be done by the end of the week. Ah, you were the volunteer :-) Recently I fixed RPM already, so we should somewhen deciede who's patch is the nicer. Ralf ^ permalink raw reply [flat|nested] 22+ messages in thread
* mipseb and mipsel patches to RPM... 1998-01-15 0:03 ` ralf @ 1998-01-15 3:56 ` Alex deVries 1998-01-15 5:25 ` ralf 0 siblings, 1 reply; 22+ messages in thread From: Alex deVries @ 1998-01-15 3:56 UTC (permalink / raw) To: ewt; +Cc: ralf, SGI Linux On Thu, 15 Jan 1998 ralf@uni-koblenz.de wrote: > Ah, you were the volunteer :-) Recently I fixed RPM already, so we should > somewhen deciede who's patch is the nicer. Erik, Feel free to apply either Ralf's or my patch (below my .sig). It supports auto-byte order detection in defaultMachine(), and patches to the rpmrc and rpm.magic files to change the name to mipseb and mipsel. I've made little-endian mips 11, and renamed the existing mips to mipseb. - Alex -- Alex deVries Run Linux on everything, System Administrator run everything on Linux. The EngSoc Project Send spam to spam@engsoc.carleton.ca. z diff -rc rpm-2.4.99.orig/lib/rpmrc.c rpm-2.4.99/lib/rpmrc.c *** rpm-2.4.99.orig/lib/rpmrc.c Tue Jan 13 16:08:24 1998 --- rpm-2.4.99/lib/rpmrc.c Wed Jan 14 19:07:28 1998 *************** *** 669,674 **** --- 669,681 ---- char * chptr; struct canonEntry * canon; + #if defined (__mips) + union { + long l; + char c[sizeof (long)]; + } u; + #endif + if (!gotDefaults) { uname(&un); if (!strcmp(un.sysname, "AIX")) { *************** *** 681,686 **** --- 688,702 ---- chptr = un.machine; while (*chptr++) if (*chptr == '/') *chptr = '-'; + + #if defined(__mips) + u.l = 1; + if (u.c[sizeof (long) - 1] == 1) { + strcpy(un.machine, "mipseb"); + } else { + strcpy(un.machine, "mipsel"); + } + #endif #if defined(__hpux) && defined(_SC_CPU_VERSION) { diff -rc rpm-2.4.99.orig/lib-rpmrc.in rpm-2.4.99/lib-rpmrc.in *** rpm-2.4.99.orig/lib-rpmrc.in Tue Jan 13 16:08:23 1998 --- rpm-2.4.99/lib-rpmrc.in Wed Jan 14 18:41:13 1998 *************** *** 41,47 **** arch_canon: sun4c: sparc 3 arch_canon: sun4d: sparc 3 # This is really a place holder for MIPS. ! arch_canon: mips: mips 4 arch_canon: ppc: ppc 5 arch_canon: m68k: m68k 6 arch_canon: IP: sgi 7 --- 41,47 ---- arch_canon: sun4c: sparc 3 arch_canon: sun4d: sparc 3 # This is really a place holder for MIPS. ! arch_canon: mipseb: mipseb 4 arch_canon: ppc: ppc 5 arch_canon: m68k: m68k 6 arch_canon: IP: sgi 7 *************** *** 50,55 **** --- 50,56 ---- arch_canon: 9000/712: hppa1.1 9 arch_canon: sun4u: usparc 10 + arch_canon: mipsel: mipsel 11 ############################################################# # Canonical OS names and numbers *************** *** 118,124 **** arch_compat: sparc: noarch arch_compat: ppc: noarch ! arch_compat: mips: noarch arch_compat: hppa1.1: hppa1.0 arch_compat: hppa1.0: noarch --- 119,126 ---- arch_compat: sparc: noarch arch_compat: ppc: noarch ! arch_compat: mipseb: noarch ! arch_compat: mipsel: noarch arch_compat: hppa1.1: hppa1.0 arch_compat: hppa1.0: noarch *************** *** 137,140 **** buildarch_compat: alpha: noarch buildarch_compat: m68k: noarch buildarch_compat: ppc: noarch ! buildarch_compat: mips: noarch --- 139,144 ---- buildarch_compat: alpha: noarch buildarch_compat: m68k: noarch buildarch_compat: ppc: noarch ! buildarch_compat: mipsel: noarch ! buildarch_compat: mipseb: noarch ! diff -rc rpm-2.4.99.orig/rpm.magic rpm-2.4.99/rpm.magic *** rpm-2.4.99.orig/rpm.magic Tue Jan 13 16:08:23 1998 --- rpm-2.4.99/rpm.magic Wed Jan 14 16:43:43 1998 *************** *** 10,17 **** >>8 beshort 1 i386 >>8 beshort 2 Alpha >>8 beshort 3 Sparc ! >>8 beshort 4 MIPS >>8 beshort 5 PowerPC >>8 beshort 6 68000 >>8 beshort 7 SGI >>10 string x %s --- 10,18 ---- >>8 beshort 1 i386 >>8 beshort 2 Alpha >>8 beshort 3 Sparc ! >>8 beshort 4 MIPS big endian >>8 beshort 5 PowerPC >>8 beshort 6 68000 >>8 beshort 7 SGI + >>8 beshort 11 MIPS little endian >>10 string x %s ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: mipseb and mipsel patches to RPM... 1998-01-15 3:56 ` mipseb and mipsel patches to RPM Alex deVries @ 1998-01-15 5:25 ` ralf 1998-01-15 18:59 ` Alex deVries 0 siblings, 1 reply; 22+ messages in thread From: ralf @ 1998-01-15 5:25 UTC (permalink / raw) To: Alex deVries; +Cc: ewt, SGI Linux On Wed, Jan 14, 1998 at 10:56:58PM -0500, Alex deVries wrote: You can solve the byteorder detection issue more efficient at compile time by something like: #if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) /* little endian */ strcpy(un.machine, "mipsel"); #else #if defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) /* big endian */ strcpy(un.machine, "mipseb"); #else /* This doesn't happen (TM) */ #endif #endif Other than that my patch is almost down to the byte the same. Ralf > diff -rc rpm-2.4.99.orig/lib/rpmrc.c rpm-2.4.99/lib/rpmrc.c > *** rpm-2.4.99.orig/lib/rpmrc.c Tue Jan 13 16:08:24 1998 > --- rpm-2.4.99/lib/rpmrc.c Wed Jan 14 19:07:28 1998 > *************** > *** 669,674 **** > --- 669,681 ---- > char * chptr; > struct canonEntry * canon; > > + #if defined (__mips) > + union { > + long l; > + char c[sizeof (long)]; > + } u; > + #endif > + > if (!gotDefaults) { > uname(&un); > if (!strcmp(un.sysname, "AIX")) { > *************** > *** 681,686 **** > --- 688,702 ---- > chptr = un.machine; > while (*chptr++) > if (*chptr == '/') *chptr = '-'; > + > + #if defined(__mips) > + u.l = 1; > + if (u.c[sizeof (long) - 1] == 1) { > + strcpy(un.machine, "mipseb"); > + } else { > + strcpy(un.machine, "mipsel"); > + } > + #endif > > #if defined(__hpux) && defined(_SC_CPU_VERSION) > { > diff -rc rpm-2.4.99.orig/lib-rpmrc.in rpm-2.4.99/lib-rpmrc.in > *** rpm-2.4.99.orig/lib-rpmrc.in Tue Jan 13 16:08:23 1998 > --- rpm-2.4.99/lib-rpmrc.in Wed Jan 14 18:41:13 1998 > *************** > *** 41,47 **** > arch_canon: sun4c: sparc 3 > arch_canon: sun4d: sparc 3 > # This is really a place holder for MIPS. > ! arch_canon: mips: mips 4 > arch_canon: ppc: ppc 5 > arch_canon: m68k: m68k 6 > arch_canon: IP: sgi 7 > --- 41,47 ---- > arch_canon: sun4c: sparc 3 > arch_canon: sun4d: sparc 3 > # This is really a place holder for MIPS. > ! arch_canon: mipseb: mipseb 4 > arch_canon: ppc: ppc 5 > arch_canon: m68k: m68k 6 > arch_canon: IP: sgi 7 > *************** > *** 50,55 **** > --- 50,56 ---- > arch_canon: 9000/712: hppa1.1 9 > > arch_canon: sun4u: usparc 10 > + arch_canon: mipsel: mipsel 11 > > ############################################################# > # Canonical OS names and numbers > *************** > *** 118,124 **** > arch_compat: sparc: noarch > > arch_compat: ppc: noarch > ! arch_compat: mips: noarch > > arch_compat: hppa1.1: hppa1.0 > arch_compat: hppa1.0: noarch > --- 119,126 ---- > arch_compat: sparc: noarch > > arch_compat: ppc: noarch > ! arch_compat: mipseb: noarch > ! arch_compat: mipsel: noarch > > arch_compat: hppa1.1: hppa1.0 > arch_compat: hppa1.0: noarch > *************** > *** 137,140 **** > buildarch_compat: alpha: noarch > buildarch_compat: m68k: noarch > buildarch_compat: ppc: noarch > ! buildarch_compat: mips: noarch > --- 139,144 ---- > buildarch_compat: alpha: noarch > buildarch_compat: m68k: noarch > buildarch_compat: ppc: noarch > ! buildarch_compat: mipsel: noarch > ! buildarch_compat: mipseb: noarch > ! > diff -rc rpm-2.4.99.orig/rpm.magic rpm-2.4.99/rpm.magic > *** rpm-2.4.99.orig/rpm.magic Tue Jan 13 16:08:23 1998 > --- rpm-2.4.99/rpm.magic Wed Jan 14 16:43:43 1998 > *************** > *** 10,17 **** > >>8 beshort 1 i386 > >>8 beshort 2 Alpha > >>8 beshort 3 Sparc > ! >>8 beshort 4 MIPS > >>8 beshort 5 PowerPC > >>8 beshort 6 68000 > >>8 beshort 7 SGI > >>10 string x %s > --- 10,18 ---- > >>8 beshort 1 i386 > >>8 beshort 2 Alpha > >>8 beshort 3 Sparc > ! >>8 beshort 4 MIPS big endian > >>8 beshort 5 PowerPC > >>8 beshort 6 68000 > >>8 beshort 7 SGI > + >>8 beshort 11 MIPS little endian > >>10 string x %s ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: mipseb and mipsel patches to RPM... 1998-01-15 5:25 ` ralf @ 1998-01-15 18:59 ` Alex deVries 1998-01-15 19:44 ` Erik Troan 1998-01-15 19:51 ` Erik Troan 0 siblings, 2 replies; 22+ messages in thread From: Alex deVries @ 1998-01-15 18:59 UTC (permalink / raw) To: ewt; +Cc: ralf, SGI Linux Erik, Please consider this as the final patch for the mips byteorder issue. It'd be nice to get this into the next version of rpm (101?). Thanks! - Alex diff -rc rpm-2.4.99.orig/lib/rpmrc.c rpm-2.4.99/lib/rpmrc.c *** rpm-2.4.99.orig/lib/rpmrc.c Tue Jan 13 16:08:24 1998 --- rpm-2.4.99/lib/rpmrc.c Thu Jan 15 12:21:29 1998 *************** *** 682,687 **** --- 682,699 ---- while (*chptr++) if (*chptr == '/') *chptr = '-'; + #if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) + /* little endian */ + strcpy(un.machine, "mipsel"); + #else + #if defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) + /* big endian */ + strcpy(un.machine, "mipseb"); + #else + /* This doesn't happen (TM) */ + #endif + #endif + #if defined(__hpux) && defined(_SC_CPU_VERSION) { int cpu_version = sysconf(_SC_CPU_VERSION); diff -rc rpm-2.4.99.orig/lib-rpmrc.in rpm-2.4.99/lib-rpmrc.in *** rpm-2.4.99.orig/lib-rpmrc.in Tue Jan 13 16:08:23 1998 --- rpm-2.4.99/lib-rpmrc.in Wed Jan 14 18:41:13 1998 *************** *** 41,47 **** arch_canon: sun4c: sparc 3 arch_canon: sun4d: sparc 3 # This is really a place holder for MIPS. ! arch_canon: mips: mips 4 arch_canon: ppc: ppc 5 arch_canon: m68k: m68k 6 arch_canon: IP: sgi 7 --- 41,47 ---- arch_canon: sun4c: sparc 3 arch_canon: sun4d: sparc 3 # This is really a place holder for MIPS. ! arch_canon: mipseb: mipseb 4 arch_canon: ppc: ppc 5 arch_canon: m68k: m68k 6 arch_canon: IP: sgi 7 *************** *** 50,55 **** --- 50,56 ---- arch_canon: 9000/712: hppa1.1 9 arch_canon: sun4u: usparc 10 + arch_canon: mipsel: mipsel 11 ############################################################# # Canonical OS names and numbers *************** *** 118,124 **** arch_compat: sparc: noarch arch_compat: ppc: noarch ! arch_compat: mips: noarch arch_compat: hppa1.1: hppa1.0 arch_compat: hppa1.0: noarch --- 119,126 ---- arch_compat: sparc: noarch arch_compat: ppc: noarch ! arch_compat: mipseb: noarch ! arch_compat: mipsel: noarch arch_compat: hppa1.1: hppa1.0 arch_compat: hppa1.0: noarch *************** *** 137,140 **** buildarch_compat: alpha: noarch buildarch_compat: m68k: noarch buildarch_compat: ppc: noarch ! buildarch_compat: mips: noarch --- 139,144 ---- buildarch_compat: alpha: noarch buildarch_compat: m68k: noarch buildarch_compat: ppc: noarch ! buildarch_compat: mipsel: noarch ! buildarch_compat: mipseb: noarch ! diff -rc rpm-2.4.99.orig/rpm.magic rpm-2.4.99/rpm.magic *** rpm-2.4.99.orig/rpm.magic Tue Jan 13 16:08:23 1998 --- rpm-2.4.99/rpm.magic Wed Jan 14 16:43:43 1998 *************** *** 10,17 **** >>8 beshort 1 i386 >>8 beshort 2 Alpha >>8 beshort 3 Sparc ! >>8 beshort 4 MIPS >>8 beshort 5 PowerPC >>8 beshort 6 68000 >>8 beshort 7 SGI >>10 string x %s --- 10,18 ---- >>8 beshort 1 i386 >>8 beshort 2 Alpha >>8 beshort 3 Sparc ! >>8 beshort 4 MIPS big endian >>8 beshort 5 PowerPC >>8 beshort 6 68000 >>8 beshort 7 SGI + >>8 beshort 11 MIPS little endian >>10 string x %s -- Alex deVries Run Linux on everything, System Administrator run everything on Linux. The EngSoc Project Send spam to spam@engsoc.carleton.ca. ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: mipseb and mipsel patches to RPM... 1998-01-15 18:59 ` Alex deVries @ 1998-01-15 19:44 ` Erik Troan 1998-01-15 19:51 ` Erik Troan 1 sibling, 0 replies; 22+ messages in thread From: Erik Troan @ 1998-01-15 19:44 UTC (permalink / raw) To: Alex deVries; +Cc: ralf, SGI Linux On Thu, 15 Jan 1998, Alex deVries wrote: > Please consider this as the final patch for the mips byteorder > issue. It'd be nice to get this into the next version of rpm (101?). I'll get this in the next version. In the guture, please use unified diffs though -- they tend to be easier to deal with. Erik ------------------------------------------------------------------------------- | "For the next two hours, VH1 will be filled with foul-mouthed, | | crossdressing Australians. Viewer discretion is advised." | | | | Erik Troan = ewt@redhat.com = ewt@sunsite.unc.edu | ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: mipseb and mipsel patches to RPM... 1998-01-15 18:59 ` Alex deVries 1998-01-15 19:44 ` Erik Troan @ 1998-01-15 19:51 ` Erik Troan 1 sibling, 0 replies; 22+ messages in thread From: Erik Troan @ 1998-01-15 19:51 UTC (permalink / raw) To: Alex deVries; +Cc: ralf, SGI Linux On Thu, 15 Jan 1998, Alex deVries wrote: > Please consider this as the final patch for the mips byteorder > issue. It'd be nice to get this into the next version of rpm (101?). Here's the final patch I used. It's basically the same as yours, I just changed the #if stuff to look a bit cleaner. I also made the old 'mips' architecture compatible with your new 'mipseb' one -- that should save you some migration headaches. Let me know if this doesn't look right. Erik Index: lib-rpmrc.in =================================================================== RCS file: /usr/rhs/CVS/rpm/lib-rpmrc.in,v retrieving revision 1.32 diff -u -r1.32 lib-rpmrc.in --- lib-rpmrc.in 1997/12/30 19:26:15 1.32 +++ lib-rpmrc.in 1998/01/15 19:51:02 @@ -41,7 +41,7 @@ arch_canon: sun4c: sparc 3 arch_canon: sun4d: sparc 3 # This is really a place holder for MIPS. -arch_canon: mips: mips 4 +arch_canon: mipseb: mipseb 4 arch_canon: ppc: ppc 5 arch_canon: m68k: m68k 6 arch_canon: IP: sgi 7 @@ -50,6 +50,7 @@ arch_canon: 9000/712: hppa1.1 9 arch_canon: sun4u: usparc 10 +arch_canon: mipsel: mipsel 11 ############################################################# # Canonical OS names and numbers @@ -118,10 +119,13 @@ arch_compat: sparc: noarch arch_compat: ppc: noarch -arch_compat: mips: noarch +arch_compat: mipseb: noarch +arch_compat: mipsel: noarch arch_compat: hppa1.1: hppa1.0 arch_compat: hppa1.0: noarch +# we used to call mipseb just mips -- let those packages still work +arch_compat: mipseb: mips os_compat: IRIX64: IRIX @@ -137,4 +141,6 @@ buildarch_compat: alpha: noarch buildarch_compat: m68k: noarch buildarch_compat: ppc: noarch -buildarch_compat: mips: noarch +buildarch_compat: mipsel: noarch +buildarch_compat: mipseb: noarch + Index: rpm.magic =================================================================== RCS file: /usr/rhs/CVS/rpm/rpm.magic,v retrieving revision 2.2 diff -u -r2.2 rpm.magic --- rpm.magic 1996/05/22 20:39:33 2.2 +++ rpm.magic 1998/01/15 19:51:02 @@ -10,8 +10,9 @@ >>8 beshort 1 i386 >>8 beshort 2 Alpha >>8 beshort 3 Sparc ->>8 beshort 4 MIPS +>>8 beshort 4 MIPS big endian >>8 beshort 5 PowerPC >>8 beshort 6 68000 >>8 beshort 7 SGI +>>8 beshort 11 MIPS little endian >>10 string x %s Index: lib/rpmrc.c =================================================================== RCS file: /usr/rhs/CVS/rpm/lib/rpmrc.c,v retrieving revision 2.57 diff -u -r2.57 rpmrc.c --- rpmrc.c 1998/01/12 16:58:16 2.57 +++ rpmrc.c 1998/01/15 19:51:02 @@ -682,6 +682,14 @@ while (*chptr++) if (*chptr == '/') *chptr = '-'; + #if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) + /* little endian */ + strcpy(un.machine, "mipsel"); + #elif defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) + /* big endian */ + strcpy(un.machine, "mipseb"); + #endif + #if defined(__hpux) && defined(_SC_CPU_VERSION) { int cpu_version = sysconf(_SC_CPU_VERSION); ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 20:58 ` Thomas Bogendoerfer 1998-01-14 22:17 ` Alan Cox @ 1998-01-14 23:59 ` Miguel de Icaza 1998-01-15 23:42 ` ralf 1 sibling, 1 reply; 22+ messages in thread From: Miguel de Icaza @ 1998-01-14 23:59 UTC (permalink / raw) To: tsbogend; +Cc: linux > hmm, to do this with only one src.rpm, we need a little support from > rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, > that for .spec execution mips is defined for bot mipsel and mipseb, because > there are changes, which work for both and we only need to seperate changes > like that needed by ncompress. Comments ? Does anybody how to do this ? The code should be fixed to autoconfigure itself by using the __LITTLE_ENDIAN and __BIG_ENDIAN macros that are defined somewhere in /usr/include. And ship the patch with this. Miguel. ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 23:59 ` The world's worst RPM Miguel de Icaza @ 1998-01-15 23:42 ` ralf 0 siblings, 0 replies; 22+ messages in thread From: ralf @ 1998-01-15 23:42 UTC (permalink / raw) To: Miguel de Icaza; +Cc: tsbogend, linux On Wed, Jan 14, 1998 at 05:59:27PM -0600, Miguel de Icaza wrote: > > hmm, to do this with only one src.rpm, we need a little support from > > rpm. At the moment mips is defined for mipsel and mipseb. I would suggest, > > that for .spec execution mips is defined for bot mipsel and mipseb, because > > there are changes, which work for both and we only need to seperate changes > > like that needed by ncompress. Comments ? Does anybody how to do this ? > > The code should be fixed to autoconfigure itself by using the > __LITTLE_ENDIAN and __BIG_ENDIAN macros that are defined somewhere in > /usr/include. And ship the patch with this. I actually have some patch ... somewhere on a disk ... Ralf ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: The world's worst RPM 1998-01-14 14:36 ` Alan Cox 1998-01-14 14:29 ` ralf @ 1998-01-14 14:36 ` Alan Cox 1 sibling, 0 replies; 22+ messages in thread From: Alan Cox @ 1998-01-14 14:36 UTC (permalink / raw) To: Oliver Frommel; +Cc: adevries, linux I've stuck some more new rpms on ftp.uk.linux.org incoming and SGI dir including sox, xmorph, ncompress, sysklogd Alan ^ permalink raw reply [flat|nested] 22+ messages in thread
end of thread, other threads:[~1998-01-16 0:02 UTC | newest] Thread overview: 22+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 1998-01-14 6:43 The world's worst RPM Alex deVries 1998-01-14 10:28 ` ralf 1998-01-14 16:38 ` Alex deVries 1998-01-14 13:28 ` Oliver Frommel 1998-01-14 13:58 ` ralf 1998-01-14 16:42 ` Alex deVries 1998-01-14 14:36 ` Alan Cox 1998-01-14 14:29 ` ralf 1998-01-14 15:07 ` Alan Cox 1998-01-14 20:58 ` Thomas Bogendoerfer 1998-01-14 22:17 ` Alan Cox 1998-01-14 22:17 ` Alan Cox 1998-01-14 22:25 ` Alex deVries 1998-01-15 0:03 ` ralf 1998-01-15 3:56 ` mipseb and mipsel patches to RPM Alex deVries 1998-01-15 5:25 ` ralf 1998-01-15 18:59 ` Alex deVries 1998-01-15 19:44 ` Erik Troan 1998-01-15 19:51 ` Erik Troan 1998-01-14 23:59 ` The world's worst RPM Miguel de Icaza 1998-01-15 23:42 ` ralf 1998-01-14 14:36 ` Alan Cox
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox