2013/8/4, José Vázquez <ppvazquez...@gmail.com>: > 2013/8/1, José Vázquez <ppvazquez...@gmail.com>: >> We didn't discarded distro or PC problem due to the different >> errors we noticed. >> In this moment i have no idea how to confirm if this issues are >> caused by the PC (Debian 6 in my case) or an errata in the BCM6368. >> VR-3025u wiki: >> http://wiki.openwrt.org/toh/comtrend/vr3025u >> VR-3025un wiki: >> http://wiki.openwrt.org/toh/comtrend/vr3025un >> > This time made the compilation with trunk r36114 (before bcm63xx > smp) and no problems enabling FPU emulation in the kernel and all > works as expected. > http://pastebin.com/qJaNeKbm > Seems that host PC and/or silicon errata problems can be thrown > to the trashcan. Could be that MIPS fpu emulation is loaded and > enabled before the second core initialization? How i can help with > this issue? > > Regards: > > José Vázquez > Finally, my colleague and me tested trunk r37737 enabling FPU emulation and switching to kernel 3.10.4 and now we haven't noticed nothing strange with jffs2. Jonas Gorski told us that would be a good idea make a compilation with GCC 4.8-linaro and the results were the same: all works fine. To summarize: the issue in jffs2 caused a rename in files which were edited or created once the router was rebooted or powercycled, and only with bcm6368 SoCs (we haven't boards with bcm6362) and CONFIG_MIPS_FPU_EMU=y selected in the kernel.
A lot of thanks to the OpenWRT team for their advices, because without them we would still fighting with this issue. Best regards: José Vázquez P.S.: there is a repository in github based in trunk r37737 for the seguridadwireless forum users; still needs a couple of additions in b43_pci_bridge.c but works fine. https://github.com/Pteridium/OpenWRT-SMP-initial _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
