[Bf-cycles] Need testers for the Disney BSDF implementation

[Jens Verwiebe]

hres a more detailed debug info ( cpu ):

(gdb) bt
#0  0x00007ffff5ea4c37 in __GI_raise (sig=sig at entry=6) at 
../nptl/sysdeps/unix/sysv/linux/raise.c:56
#1  0x00007ffff5ea8028 in __GI_abort () at abort.c:89
#2  0x00007ffff5e9dbf6 in __assert_fail_base (fmt=0x7ffff5fee3b8 
"%s%s%s:%u: %s%sAssertion `%s' failed.\n%n",
     assertion=assertion at entry=0x82b73af "label & LABEL_TRANSMIT",
     file=file at entry=0x82b7348 
"/media/Workdata3/Development/blender_git/cycles_disney_brdf/intern/cycles/kernel/./kernel_path_state.h", 
line=line at entry=100,
     function=function at entry=0x82b93e0 "void 
ccl::path_state_next(ccl::KernelGlobals*, ccl::PathState*, int)") at 
assert.c:92
#3  0x00007ffff5e9dca2 in __GI___assert_fail (assertion=0x82b73af "label 
& LABEL_TRANSMIT",
     file=0x82b7348 
"/media/Workdata3/Development/blender_git/cycles_disney_brdf/intern/cycles/kernel/./kernel_path_state.h", 

     line=100, function=0x82b93e0 "void 
ccl::path_state_next(ccl::KernelGlobals*, ccl::PathState*, int)") at 
assert.c:101
#4  0x0000000004cdda4a in ?? ()
#5  0x0000000004d0e85f in 
ccl::kernel_cpu_avx2_path_trace(ccl::KernelGlobals*, float*, unsigned 
int*, int, int, int, int, int) ()
#6  0x0000000003e38124 in 
ccl::CPUDevice::thread_path_trace(ccl::DeviceTask&) ()
#7  0x0000000003e37afa in ccl::CPUDevice::thread_run(ccl::DeviceTask*) ()
#8  0x0000000003e39fd6 in ?? ()
#9  0x0000000003e39ecd in ?? ()
#10 0x0000000003e39c8e in ?? ()
#11 0x0000000003e39bcb in ?? ()
#12 0x0000000004e97f34 in ?? ()
#13 0x0000000004e96efe in ccl::TaskScheduler::thread_run(int) ()


Am 20.07.2016 um 15:13 schrieb "Pascal Schön":
> Hey Jens,
> thanks for your reply and your work. I've done a Windows 64bit build 
> that can be downloaded here (thanks to Bastien for your 
> advice!): https://www.dropbox.com/s/rkpbwr18dwptarr/blender_cycles_disney_brdf_win64_latest.zip?dl=1
> To the issue with SSS:
> I've encountered a similar error but can't reproduce it. When using 
> CUDA, I get the message that I'm out of memory but I also get this 
> error when using the default SSS shader. On the CPU it once crashed 
> with an EXCEPTION_ACCESS_VIOLATION. But since then no more crashes 
> happened. I have also tested it with a new scene and with all other 
> parameters on and off.
> Maybe someone can figure out what's happening here.
> Pascal
> *Gesendet:* Mittwoch, 20. Juli 2016 um 14:15 Uhr
> *Von:* "Jens Verwiebe" <info at jensverwiebe.de>
> *An:* bf-cycles at blender.org
> *Betreff:* [Bf-cycles] Need testers for the Disney BSDF implementation
> Hi Pascal
>
> I made a Linux portable build ( all distro ) of cycles_disney_brdf
>
> -- PACKAGE-INFORMATION --
> Branch: cycles_disney_brdf
> Revision: dbad91c
> Submodules: locale 4c06104 addons 2155eb5 addons_contrib 98f1144
> OS: GNU/Linux, Architecture: x86_64, GLIBC: 2.19
> Builddate: Mi 20. Jul 11:46:49 UTC 2016
> Filesize: 72398292 byte
> Shasum: 5d535d7b5b6f12a6bf86c021fc530233586a1b0d
> URL:
> http://www.jensverwiebe.de/Blender/blender_cycles_disney_brdf_linux64_latest.tar.xz
>
> Initial testing works so far, but there is an issue in SSS:
>
> In cuda:
> CUDA error: Illegal address in cuCtxSynchronize()
>
> In CPU also just getting a segfault ( non debug build for now )
>
> Cheers ... Jens
>
> --
>
> Jens Verwiebe
> Allerskehre 44 - 22309 Hamburg
>
> Tel.: +49 40 68 78 50
> mobile: +49 172 400 49 07
> mailto: info at jensverwiebe.de
> web: http://www.jensverwiebe.de
>
> _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> https://lists.blender.org/mailman/listinfo/bf-cycles
>
>
> _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> https://lists.blender.org/mailman/listinfo/bf-cycles

-- 

Jens Verwiebe
Allerskehre 44 - 22309 Hamburg

Tel.: +49 40 68 78 50
mobile: +49 172 400 49 07
mailto: info at jensverwiebe.de
web: http://www.jensverwiebe.de

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/bf-cycles/attachments/20160720/1dbe9c70/attachment.htm