Menu
▾
▴
devil-linux-develop
|
From: Bruce S. <bw...@ar...> - 2003-08-09 23:41:32
|
# cvs_sf update -d -P # ./update_src # make mrproper unpack build install iso ... patching file drivers/net/sha_locl.h patching file include/linux/ppp-comp.h patching file include/openssl/opensslconf.h patching file include/openssl/opensslv.h patching file include/openssl/rc4.h patching file include/openssl/sha.h executing linux with option build (in /data/build/tmp/linux-2.4.21) applying patch: /data/build/tmp/kernel-patches/cflags_extra.diff patching file Makefile applying patch: /data/build/tmp/kernel-patches/cramfs-initrd.diff patching file init/do_mounts.c applying patch: /data/build/tmp/kernel-patches/patch-watchdogdriver-pcwd.diff patching file drivers/char/pcwd.c make: *** empty string invalid as file name. Stop. make: *** empty string invalid as file name. Stop. ERROR /data/build/scripts/linux build failed make: *** [build] Error 1 |
|
From: Heiko Z. <he...@zu...> - 2003-08-10 23:40:27
|
Hey, how was the vacation? We did so many changes, so you start with a clean lfssystem. The patch where it fails (pcwd driver) didn't get modified for a while, so it should work fine. cya Heiko Bruce Smith wrote: ># cvs_sf update -d -P ># ./update_src ># make mrproper unpack build install iso >... >patching file drivers/net/sha_locl.h >patching file include/linux/ppp-comp.h >patching file include/openssl/opensslconf.h >patching file include/openssl/opensslv.h >patching file include/openssl/rc4.h >patching file include/openssl/sha.h > >executing linux with option build (in /data/build/tmp/linux-2.4.21) >applying patch: /data/build/tmp/kernel-patches/cflags_extra.diff >patching file Makefile >applying patch: /data/build/tmp/kernel-patches/cramfs-initrd.diff >patching file init/do_mounts.c >applying patch: >/data/build/tmp/kernel-patches/patch-watchdogdriver-pcwd.diff >patching file drivers/char/pcwd.c >make: *** empty string invalid as file name. Stop. >make: *** empty string invalid as file name. Stop. >ERROR >/data/build/scripts/linux build failed >make: *** [build] Error 1 > > |
|
From: Bruce S. <bw...@ar...> - 2003-08-11 12:38:27
|
> how was the vacation?
Great! Love 'da UP, eh! (sorry, local humor :)
> We did so many changes, so you start with a clean lfssystem.
> The patch where it fails (pcwd driver) didn't get modified for a while,
> so it should work fine.
OK, I started with a clean lfssystem, and it got a little farther:
gcc -L../../../lib -Wl,-rpath -Wl,/usr/lib -g -O2 -Wall
-Wmissing-prototypes -Wcast-qual -Wcast-align -Wconversion -Wshadow
-Wno-comment -pedantic -o client client.o rpc_test_clnt.o \
-lgssrpc -lgssapi_krb5 -lkrb5 -lk5crypto -lcom_err -lresolv
/usr/bin/ld: client: hidden symbol `__stack_smash_handler' in
/usr/lib/gcc-lib/i586-pc-linux-gnu/3.2.3/libgcc.a(_stack_smash_handler.oS) is referenced by DSO
collect2: ld returned 1 exit status
make[4]: *** [client] Error 1
make[4]: Leaving directory
`/data/build/tmp/krb5-1.3/src/lib/rpc/unit-test'
make[3]: *** [all-recurse] Error 1
make[3]: Leaving directory `/data/build/tmp/krb5-1.3/src/lib/rpc'
make[2]: *** [all-recurse] Error 1
make[2]: Leaving directory `/data/build/tmp/krb5-1.3/src/lib'
make[1]: *** [all-recurse] Error 1
make[1]: Leaving directory `/data/build/tmp/krb5-1.3/src'
ERROR
/data/build/scripts/krb5 build failed
make: *** [build] Error 1
|
|
From: Heiko Z. <he...@zu...> - 2003-08-11 13:40:32
|
Bruce Smith wrote: >>how was the vacation? > > > Great! Love 'da UP, eh! (sorry, local humor :) > > >>We did so many changes, so you start with a clean lfssystem. >>The patch where it fails (pcwd driver) didn't get modified for a while, >>so it should work fine. > > > OK, I started with a clean lfssystem, and it got a little farther: That error is supposed to happen. Kerberos doesn't compile at the moment with SSP enabled. Just disable it in menuconfig and it should run through. Heiko |
|
From: Bruce S. <bw...@ar...> - 2003-08-14 18:27:50
|
> Kerberos doesn't compile at the moment with SSP enabled. > Just disable it in menuconfig and it should run through. What would happen if we don't turn on the SSP option in krb5 but leave it on in all the other scripts? Would that cause other problems with Kerberos? I won't do it until I hear back. - BS |
|
From: Manu <ee...@fr...> - 2003-08-14 19:56:21
|
Hi, Bruce Smith a écrit: >>Kerberos doesn't compile at the moment with SSP enabled. >>Just disable it in menuconfig and it should run through. >> >> > >What would happen if we don't turn on the SSP option in krb5 >but leave it on in all the other scripts? Would that cause >other problems with Kerberos? I won't do it until I hear back. > > Don't really know but it's a thing to try ... kerberos won't be protected by ssp, that's all just include -fno-stack-protector in the compiler options for Kerberos package and see what happen ... Cya MaNU > - BS > > > > >------------------------------------------------------- >This SF.Net email sponsored by: Free pre-built ASP.NET sites including >Data Reports, E-commerce, Portals, and Forums are available now. >Download today and enter to win an XBOX or Visual Studio .NET. >http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 >_______________________________________________ >Devil-linux-develop mailing list >Dev...@li... >https://lists.sourceforge.net/lists/listinfo/devil-linux-develop > > > > |
|
From: Bruce S. <bw...@ar...> - 2003-08-14 21:09:12
|
> > > Kerberos doesn't compile at the moment with SSP enabled. > > > Just disable it in menuconfig and it should run through. > > > > What would happen if we don't turn on the SSP option in krb5 > > but leave it on in all the other scripts? Would that cause > > other problems with Kerberos? I won't do it until I hear back. > > > Don't really know but it's a thing to try ... kerberos won't be > protected by ssp, that's all > just include -fno-stack-protector in the compiler options for Kerberos > package and see what happen ... I tried it both with and without that option, and it won't compile. Maybe because GCC or the kernel (or something) was compiled with SSP? Oh well, it was worth a try. Never mind! :-) - BS |
|
From: Heiko Z. <he...@zu...> - 2003-08-18 22:55:29
|
Bruce Smith wrote: >>>>Kerberos doesn't compile at the moment with SSP enabled. >>>>Just disable it in menuconfig and it should run through. >>> >>>What would happen if we don't turn on the SSP option in krb5 >>>but leave it on in all the other scripts? Would that cause >>>other problems with Kerberos? I won't do it until I hear back. >>> >> >>Don't really know but it's a thing to try ... kerberos won't be >>protected by ssp, that's all >>just include -fno-stack-protector in the compiler options for Kerberos >>package and see what happen ... > > > I tried it both with and without that option, and it won't compile. > Maybe because GCC or the kernel (or something) was compiled with SSP? > > Oh well, it was worth a try. Never mind! :-) I'll work some more on that problem, when I'm in the mood for it. ;-) Heiko |
|
From: Heiko Z. <he...@zu...> - 2003-08-18 23:35:30
|
Heiko Zuerker wrote: > Bruce Smith wrote: > >>>>> Kerberos doesn't compile at the moment with SSP enabled. >>>>> Just disable it in menuconfig and it should run through. >>>> >>>> >>>> What would happen if we don't turn on the SSP option in krb5 >>>> but leave it on in all the other scripts? Would that cause >>>> other problems with Kerberos? I won't do it until I hear back. >>>> >>> >>> >>> Don't really know but it's a thing to try ... kerberos won't be >>> protected by ssp, that's all >>> just include -fno-stack-protector in the compiler options for Kerberos >>> package and see what happen ... >> >> >> >> I tried it both with and without that option, and it won't compile. >> Maybe because GCC or the kernel (or something) was compiled with SSP? >> >> Oh well, it was worth a try. Never mind! :-) > > > I'll work some more on that problem, when I'm in the mood for it. ;-) Kerberos 5 is fixed. Note: SSP is not enabled for the kerberos 5 package. cya Heiko |
|
From: Bruce S. <bw...@ar...> - 2003-08-11 13:42:26
|
> OK, I started with a clean lfssystem, and it got a little farther:
Actually, I spoke too soon. It did NOT get any farther.
The other abort is from a different PC (also with a new lfssystem) with
a limited menuconfig selection. I got the same error on the PC with
everything selected:
patching file pppd/sha1dgst.c
patching file pppd/sha_locl.h
patching file drivers/net/Makefile
Hunk #1 succeeded at 144 with fuzz 2 (offset 17 lines).
patching file drivers/net/md32_common.h
patching file drivers/net/mppe.h
patching file drivers/net/ppp_generic.c
Hunk #1 succeeded at 1029 (offset 23 lines).
patching file drivers/net/ppp_mppe.c
patching file drivers/net/rc4_enc.c
patching file drivers/net/rc4_locl.h
patching file drivers/net/rc4_skey.c
patching file drivers/net/sha1dgst.c
patching file drivers/net/sha_locl.h
patching file include/linux/ppp-comp.h
patching file include/openssl/opensslconf.h
patching file include/openssl/opensslv.h
patching file include/openssl/rc4.h
patching file include/openssl/sha.h
executing linux with option build (in /data/build/tmp/linux-2.4.21)
applying patch: /data/build/tmp/kernel-patches/cflags_extra.diff
patching file Makefile
applying patch: /data/build/tmp/kernel-patches/cramfs-initrd.diff
patching file init/do_mounts.c
applying patch:
/data/build/tmp/kernel-patches/patch-watchdogdriver-pcwd.diff
patching file drivers/char/pcwd.c
make: *** empty string invalid as file name. Stop.
make: *** empty string invalid as file name. Stop.
ERROR
/data/build/scripts/linux build failed
make: *** [build] Error 1
-----------------------------------------------------
It looks like there is at least two different aborts.
Did I mention both systems have new lfssystems? ;-)
- BS
|
|
From: Bruce S. <bw...@ar...> - 2003-08-11 14:31:22
|
> make: *** empty string invalid as file name. Stop. > make: *** empty string invalid as file name. Stop. > ERROR > /data/build/scripts/linux build failed > make: *** [build] Error 1 Fixed, uploaded to CVS. You cannot have quotes around $FLAGS in the make statements, when $FLAGS is null. - BS |
|
From: Heiko Z. <he...@zu...> - 2003-08-11 14:40:29
|
Bruce Smith wrote: >>make: *** empty string invalid as file name. Stop. >>make: *** empty string invalid as file name. Stop. >>ERROR >>/data/build/scripts/linux build failed >>make: *** [build] Error 1 > > > Fixed, uploaded to CVS. > > You cannot have quotes around $FLAGS in the make statements, > when $FLAGS is null. I guess you don't use the SSP, correct? There could be a few more places where it is this way, but some of them need it. cya Heiko |
|
From: Bruce S. <bw...@ar...> - 2003-08-11 14:50:42
|
> > You cannot have quotes around $FLAGS in the make statements, > > when $FLAGS is null. > > I guess you don't use the SSP, correct? I didn't rerun menuconfig on the one box where it aborted, and since SSP is a new option, it wasn't selected. > There could be a few more places where it is this way, but some of them > need it. I'll finish this run with the current options and see if it aborts in any other places. - BS |
|
From: Bruce S. <bw...@ar...> - 2003-08-11 19:33:01
|
> > There could be a few more places where it is this way, but some of them > > need it. > > I'll finish this run with the current options and see if it aborts in > any other places. I had another abort on my limited selection config. Snort aborted because it couldn't find MySQL (not selected). I selected & compiled MySQL, restarted make build, and it finished. Why is MySQL required for Snort? |
|
From: Bruce S. <bw...@ar...> - 2003-08-11 20:41:34
|
> Why is MySQL required for Snort? Because --with-mysql is specified on the ./configure line, duh! ;-) I modified the script so it only specifies that flag if MySQL is selected in the make menuconfig. I also added a --with-postgresql if PG is selected in menuconfig. Plus I checked OPENSSL_BIN and conditionally specified --with-ssl (although that appears to no longer be a configure option) - BS |
|
From: Heiko Z. <he...@zu...> - 2003-08-11 21:09:36
|
Bruce Smith wrote: >>>There could be a few more places where it is this way, but some of them >>>need it. >> >>I'll finish this run with the current options and see if it aborts in >>any other places. > > > I had another abort on my limited selection config. > > Snort aborted because it couldn't find MySQL (not selected). > I selected & compiled MySQL, restarted make build, and it finished. > > Why is MySQL required for Snort? You can modify the script if you want, so it's optional. It is used for logging into a MySQL DB, which can be used by most Snort Frontends. Heiko |
|
From: Heiko Z. <he...@zu...> - 2003-08-11 14:25:14
|
Bruce Smith wrote: >>OK, I started with a clean lfssystem, and it got a little farther: > > > Actually, I spoke too soon. It did NOT get any farther. > The other abort is from a different PC (also with a new lfssystem) with > a limited menuconfig selection. I got the same error on the PC with > everything selected: > I just verified a few things and didn't find anything. Can you send me you config, then I start a test compile here (from scratch of course). Heiko |
|
From: Bruce S. <bw...@ar...> - 2003-08-11 14:36:39
|
> I just verified a few things and didn't find anything. > Can you send me you config, then I start a test compile here (from > scratch of course). If you want to see it abort, turn off: CONFIG_GCC_STACK_PROTECTOR If you want to see it work w/stack_protector, update from CVS. :-) - BS |
