[patch] firefox35/firefox3 on amd64 security/nss/lib/freebl

[Martin Wilke]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Sun, Aug 16, 2009 at 10:06:26PM +0200, Andreas Tobler wrote:
> Hi,
> 
> the attached patch enables the -D_X86_ switch on amd64 platforms. The 
> code inside this directory mentions to use this switch on Intel procs. 
> But the filter in the Makefile only recognizes x86 and x86_64 platforms 
> and not fbsd like amd64 platforms.
> 
> The patch should apply for both foxes, 3.0.13 and 3.5.
> 
> Unfortunately the binary does not work on amd64 here. Need to test why.
> 
> I see the also a 'bad system call' on powerpc. Another patch in the 
> pipeline for building.
> 
> It is not due to the patch, it does coredump with and without patch:
> 
> Program received signal SIGSYS, Bad system call.
> [Switching to Thread 808744b40 (LWP 100165)]
> 0x00000008049922ac in ksem_init () from /lib/libc.so.7
> (gdb) bt
> #0  0x00000008049922ac in ksem_init () from /lib/libc.so.7
> #1  0x000000080498662f in sem_init () from /lib/libc.so.7
> #2  0x000000080106be0f in nsCanvasBidiProcessor::~nsCanvasBidiProcessor ()
>     from /usr/local/lib/firefox3/libxul.so
> #3  0x000000080106be88 in nsCanvasBidiProcessor::~nsCanvasBidiProcessor ()
>     from /usr/local/lib/firefox3/libxul.so
> #4  0x0000000801060a93 in nsCanvasBidiProcessor::~nsCanvasBidiProcessor ()
>     from /usr/local/lib/firefox3/libxul.so
> #5  0x000000080106274f in nsCanvasBidiProcessor::~nsCanvasBidiProcessor ()
>     from /usr/local/lib/firefox3/libxul.so
> #6  0x00000008014ecc22 in NS_GetComponentManager_P ()
>     from /usr/local/lib/firefox3/libxul.so
> #7  0x00000008014b809d in JNIEnv_::CallStaticObjectMethod ()
>     from /usr/local/lib/firefox3/libxul.so
> #8  0x00000008014ed2bd in NS_GetComponentManager_P ()
>     from /usr/local/lib/firefox3/libxul.so
> #9  0x0000000801e54f5a in _pt_root () from /usr/local/lib/libplds4.so.1
> #10 0x0000000804b8b4b1 in pthread_getprio () from /lib/libthr.so.3
> #11 0x0000000000000000 in ?? ()
> Error accessing memory address 0x7ffffe5f4000: Bad address.
> (gdb)
> 

kldload sem solved this problem.

> 
> Danke,
> Andreas

> --- security/nss/lib/freebl/Makefile.orig	2009-08-16 21:20:19.000000000 +0200
> +++ security/nss/lib/freebl/Makefile	2009-08-16 21:20:34.000000000 +0200
> @@ -79,7 +79,7 @@
>  endif
>  # some code wants _X86_ defined for intel CPUs.
>  # coreconf does this for windows, but not for Linux, FreeBSD, etc.
> -ifeq (,$(filter-out x86 x86_64,$(CPU_ARCH)))
> +ifeq (,$(filter-out amd64 x86 x86_64,$(CPU_ARCH)))
>  ifneq (,$(filter-out WIN%,$(OS_TARGET)))
>  	DEFINES += -D_X86_
>  endif

> _______________________________________________
> freebsd-gecko at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-gecko
> To unsubscribe, send any mail to "freebsd-gecko-unsubscribe at freebsd.org"


- -- 

+-----------------------+-------------------------------+
|  PGP    : 0xB1E6FCE9  |  Jabber : miwi(at)BSDCrew.de  |
|  Skype  : splash_111  |  Mail   : miwi(at)FreeBSD.org |
+-----------------------+-------------------------------+
|	Mess with the Best, Die like the Rest!		|
+-----------------------+-------------------------------+
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.11 (FreeBSD)

iEYEARECAAYFAkqJC3EACgkQdLJIhLHm/OnMfgCgpZpayvH7Jai1UVeB+jVih6mU
LEcAoMNCKvek8m/6L9GkJ2S2jKPEyQM0
=kKPG
-----END PGP SIGNATURE-----