lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Wed, 30 May 2012 13:51:37 -0700
From:	shawn <shawnlandden@...il.com>
To:	Jason Baron <jbaron@...hat.com>
Cc:	Jim Cromie <jim.cromie@...il.com>,
	Russell King <rmk+kernel@....linux.org.uk>,
	linux-kernel <linux-kernel@...r.kernel.org>
Subject: Re: include/linux/dynamic_debug.h:111:2: error: implicit
 declaration of function ‘strstr’

On Wed, 2012-05-30 at 16:34 -0400, Jason Baron wrote: 
> On Wed, May 30, 2012 at 12:40:01PM -0700, shawn wrote:
> > summary:
> > 
> > the recently added to arm, CONFIG_KERNEL_XZ is broken because
> > arch/arm/boot/compressed/decompress.c defines _LINUX_STRING_H
> > overriding <linux/string.h> used in include/linux/dynamic_debug.h:111
> > 
> 
> If arch/arm/boot/compressed/decompress.c is going to define
> '_LINUX_STRING_H', I think it needs to provide any string functions that
> might be missing. Thus in this case, if you add a:
> 
> extern char * strstr(const char *, const char *);
> 
> right before:
> 
> #include "../../../../lib/decompress_unxz.c",
> 
> does that fix things?

Yes that worked, and I've included a patch that does that. However I am
not sure that is the best way to fix this problem. 
> 
> Thanks,
> 
> -Jason
> 
> 
> > 
> > On Wed, 2012-05-30 at 11:53 -0600, Jim Cromie wrote: 
> > > On Wed, May 30, 2012 at 8:10 AM, shawn <shawnlandden@...il.com> wrote:
> > > > On Wed, 2012-05-30 at 02:35 -0600, Jim Cromie wrote:
> > > >> On Tue, May 29, 2012 at 7:43 PM, shawn <shawnlandden@...il.com> wrote:
> > > >> > jim,
> > > >> > On Tue, 2012-05-29 at 19:11 -0600, Jim Cromie wrote:
> > > >> >> On Tue, May 29, 2012 at 5:31 PM, shawn <shawnlandden@...il.com> wrote:
> > > >> >> >  SHIPPED arch/arm/boot/compressed/lib1funcs.S
> > > >> >> > In file included from include/linux/kernel.h:23:0,
> > > >> >> >                 from
> > > >> >> > arch/arm/boot/compressed/../../../../lib/xz/xz_private.h:15,
> > > >> >> >                 from
> > > >> >> > arch/arm/boot/compressed/../../../../lib/decompress_unxz.c:145,
> > > >> >> >                 from arch/arm/boot/compressed/decompress.c:50:
> > > >> >> > include/linux/dynamic_debug.h: In function
> > > >> >> > ‘ddebug_dyndbg_module_param_cb’:
> > > >> >> > include/linux/dynamic_debug.h:111:2: error: implicit declaration of
> > > >> >> > function ‘strstr’ [-Werror=implicit-function-declaration]
> > > >> >> >
> > > >> >> > git blame shows b48420c1d3019ce8d84fb8e58f4ca86b8e3655b8 as introducing
> > > >> >> > this call
> > > >> >>
> > > >> >> that symbol should have been provided by line 100
> > > >> >>
> > > >> >>      98 #else
> > > >> >>      99
> > > >> >>     100 #include <linux/string.h>
> > > >> >>     101 #include <linux/errno.h>
> > > >> >> ...
> > > >> >>     108 static inline int ddebug_dyndbg_module_param_cb(char *param, char *val,
> > > >> >>     109                                                 const char *modname)
> > > >> >>     110 {
> > > >> >>     111         if (strstr(param, "dyndbg")) {
> > > >> >
> > > >> >> which has:
> > > >> >>
> > > >> >>      74 #ifndef __HAVE_ARCH_STRSTR
> > > >> >>      75 extern char * strstr(const char *, const char *);
> > > >> >>      76 #endif
> > > >> >> does your arch have __HAVE_ARCH_STRSTR ?
> > > >> >> if so, is strstr exported under different conditions ??
> > > >> > a git grep __HAVE_ARCH_STRSTR did not show and such defines in the arm
> > > >> > tree
> > > >>
> > > >> thats odd - you should have gotten the generic one then.
> > > >>
> > > >> >> Can you send me
> > > >> >> arch/arm/boot/compressed/decompress.i  ?
> > > >> > I didn't have that file after the failure, but
> > > >>
> > > >> nor would you with success, but you did the right thing below..
> > > >>
> > > >> > make arch/arm/boot/compressed/decompress.i
> > > >> >
> > > >> > summoned it, also including string.o from that folder (which was present
> > > >> > after the failure)
> > > >> >
> > > >> > I am compiling on amd64, using
> > > >> > export ARCH=arm
> > > >> > and
> > > >> > CONFIG_CROSS_COMPILE="arm-linux-gnueabi-"
> > > >> >
> > > >> >
> > > >> > I have had issues with the cross compiler not always being called since
> > > >> > the 3.5 merge window opened up, (ld reporting corrupted binary files and
> > > >> > gcc saying it doesn't understand arm-specific flags) and have not yet
> > > >> > compiled a post-v3.4 successful kernel. This could be a symptom of that.
> > > >> > (esp as the host arch, x86 DOES have __HAVE_ARCH_STRSTR)
> > > >> >
> > > >>
> > > >> hmm
> > > >> CONFIG_CROSS_COMPILE is new in 3.5-rc isnt it ?
> > > >> what happens if you do it the old (3.4) way ?
> > > > No, CONFIG_CROSS_COMPILE is not new, I've been doing it this way for a
> > > > few versions, before with the emdebian cross toolchain. in meuconfig it
> > > > shows up under: General Config --> Cross-compiler tool prefix
> > > 
> > > Ah yes - I see its not new (far from it)
> > > 
> > > commit 84336466011c589b6af554f2f2f1fcfa1a5c1437
> > > Author: Roland McGrath <roland@...hat.com>
> > > Date:   Mon Dec 21 16:24:06 2009 -0800
> > > 
> > >     kconfig CROSS_COMPILE option
> > > 
> > > 
> > > > This is the arm toolchain that is in the Ubuntu repos.
> > > >
> > > > http://packages.ubuntu.com/precise/gcc-4.6-arm-linux-gnueabi
> > > 
> > > I went ahead and installed the arm toolchain from kernel.org,
> > > and got what look to be errors like you mentioned:
> > > 
> > > [jimc@...ucho linux-2.6-dep]$ make
> > > make[1]: Nothing to be done for `all'.
> > > make[1]: Nothing to be done for `relocs'.
> > >   CHK     include/linux/version.h
> > >   CHK     include/generated/utsrelease.h
> > >   CC      kernel/bounds.s
> > > cc1: error: unrecognized command line option '-m64'
> > > cc1: error: unrecognized command line option '-mno-red-zone'
> > > cc1: error: unrecognized command line option '-mcmodel=kernel'
> > > cc1: error: unrecognized command line option '-maccumulate-outgoing-args'
> > > make[1]: *** [kernel/bounds.s] Error 1
> > > make: *** [prepare0] Error 2
> > > 
> > > this is with toolchain dir in the front of the path, after doing
> > > make ARCH=arm defconfig
> > > make xconfig # and adding the cross-compile value
> > > 
> > > I got much further when I kept ARCH=arm as make arg;
> > > it completed the build.  Youre getting it from the environment,
> > > which should be the same, but its worth a try...
> > > 
> > > [jimc@...ucho linux-2.6-dep]$ file vmlinux*
> > > vmlinux:   ELF 32-bit LSB executable, ARM, version 1, statically
> > > linked, not stripped
> > > vmlinux.o: ELF 32-bit LSB relocatable, ARM, version 1, not stripped
> > > 
> > > FWIW, this was an in-src build, my attempt at O=../build-arm spat at me.
> > > 
> > > 
> > > > And I have successfully built kernels with this toolchain, and the
> > > > debian native arm 4.6 compiler, at the same time, using distcc. (however
> > > > I turned both distcc and ccache off for the error I sent)
> > > >
> > > > After having the issues with 3.5, I reinstalled, well parts of the tools
> > > > chain with apt-get install --reinstall.
> > > >
> > > > What is the old way to cross-compile?
> > > 
> > > from the config help text:
> > > 
> > > Same as running 'make CROSS_COMPILE=prefix-' but stored for
> > > default make runs in this kernel build directory. You don't
> > > need to set this unless you want the configured kernel build
> > > directory to select the cross-compiler automatically.
> > > 
> > > FWIW, it sounds more fiddly than using the now-quite-mature
> > > CROSS_COMPILE config-item
> > > 
> > > 
> > > 
> > > >> What toolchain are you using ?
> > > >> can you replicate it using something here ?
> > > >>
> > > >> http://kernel.org/pub/tools/crosstool/
> > > 
> > > Since I got a build using this, you should give it a go.
> > > If it works, file a bug report with ubuntu.
> > 
> > arch/arm/boot/compressed/decompress.c
> > 
> > has
> > 
> > #define _LINUX_STRING_H_
> > 
> > which prevents <linux/string.h> from loading
> > 
> > 
> > therefore: xz_private.h cannot include <linux/kernel.h>, or
> > arch/arm/boot/compressed/decompress.c must define strstr
> > 
> > I tried to remove the #include <linux/kernel.h> 
> > from xz_private.h, and replace it with more specific includes but cannot
> > because lib/xz/xz_dec_lzma2.c uses min() which is defined in the
> > <linux/kernel.h> header
> > 
> >   CC      arch/arm/boot/compressed/decompress.o
> > In file included from include/linux/kernel.h:23:0,
> >                  from
> > arch/arm/boot/compressed/../../../../lib/xz/xz_private.h:15,
> >                  from
> > arch/arm/boot/compressed/../../../../lib/decompress_unxz.c:145,
> >                  from arch/arm/boot/compressed/decompress.c:50:
> > include/linux/dynamic_debug.h: In function
> > ‘ddebug_dyndbg_module_param_cb’:
> > include/linux/dynamic_debug.h:111:2: error: implicit declaration of
> > function ‘strstr’ [-Werror=implicit-function-declaration]
> > 
> > > 
> > > >
> > > > --
> > > > -Shawn Landden
> > > >
> > > 
> > > let me know how you get on,
> > > ( learning from others' pain is one of the values of empathy! )
> > > 
> > > Jim
> > 
> > 
> > -- 
> > -Shawn Landden
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@...r.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/


-- 
-Shawn Landden

View attachment "0001-arm-add-definition-of-strstr-to-decompress.c.patch" of type "text/x-patch" (1372 bytes)

Powered by blists - more mailing lists