Date: Mon, 2 May 2016 13:22:13 -0700 From: Mark Millard <markmi@dsl-only.net> To: Adrian Chadd <adrian@freebsd.org>, freebsd-arm <freebsd-arm@freebsd.org> Subject: RE: rpi2 snapshots not booting? Message-ID: <8B4E345E-1BB0-4D07-9730-740BA48E800F@dsl-only.net>
index | next in thread | raw e-mail
On Mon May 2 18:02:16 UTC 2016 Adrian Chadd wrote: > On 30 April 2016 at 14:34, Adrian Chadd <adrian at freebsd.org wrote: > > Hiya, > > > > I'm trying to boot the -HEAD snapshot on a rpi2. I've tried two, to no success. > > [snip] > > ping - the last two snapshots didn't boot, and so anyone trying > freebsd-head on this will have a very crappy day. > > Is anyone able to help figure out what broke? > > > -a (Any following quoted text is my copy and paste of my own local material.) I have an rpi2 running: > # uname -apKU > FreeBSD rpi2 11.0-CURRENT FreeBSD 11.0-CURRENT #21 r297769M: Sat Apr 9 22:51:03 PDT 2016 markmi@FreeBSDx64:/usr/obj/clang/arm.armv6/usr/src/sys/RPI2-NODBG arm armv6 1100105 1100105 -r297769 would be between CHECKSUM.SHA256-FreeBSD-11.0-CURRENT-arm-armv6-RPI2-20160408-r297692 and CHECKSUM.SHA256-FreeBSD-11.0-CURRENT-arm-armv6-RPI2-20160429-r298793 At the time it was a buildworld buildkernel update based on a cross-build from an amd64 context. I do remember having to update rpi.dtb on the FAT partition at some point along the way. ( https://lists.freebsd.org/pipermail/freebsd-arm/2016-April/013657.html "No usable event timer found on RPI2" ) Overall this makes me wonder if there is something more specific to the snapshots that is a problem instead of there being a general 11.0-CURRENT problem. I did (and generally do) use compiler options in src.conf to force targeting armv7a/cortex-a7 for the rpi2: > WITH_LIBSOFT= > WITH_LIBCPLUSPLUS= > WITH_BINUTILS_BOOTSTRAP= > WITH_CLANG= > WITH_CLANG_IS_CC= > WITH_CLANG_FULL= > WITH_LLDB= > WITH_CLANG_EXTRAS= > WITH_BOOT= > # > WITHOUT_LIB32= > WITHOUT_ELFTOOLCHAIN_BOOTSTRAP= > WITHOUT_CLANG_BOOTSTRAP= > WITHOUT_GCC_BOOTSTRAP= > WITHOUT_GCC= > WITHOUT_GNUCXX= > # > NO_WERROR= > MALLOC_PRODUCTION= > # > WITH_DEBUG_FILES= > # > .if ${.MAKE.LEVEL} == 0 > XCC=/usr/bin/clang -target armv6--freebsd11.0-gnueabi -march=armv7a -mcpu=cortex-a7 -mno-unaligned-access > XCXX=/usr/bin/clang++ -target armv6--freebsd11.0-gnueabi -march=armv7a -mcpu=cortex-a7 -mno-unaligned-access > XCPP=/usr/bin/clang-cpp -target armv6--freebsd11.0-gnueabi -march=armv7a -mcpu=cortex-a7 -mno-unaligned-access > .export XCC > .export XCXX > .export XCPP > .endif > # > .if ${.MAKE.LEVEL} == 0 > CC=/usr/bin/clang > CXX=/usr/bin/clang++ > CPP=/usr/bin/clang-cpp > .export CC > .export CXX > .export CPP > .endif The amd64 context was updated to 11.0-CURRENT -r297769 before the rpi2 rebuild was done. So clang was already the matching version, no clang bootstrap needed as far as I know. make.conf was empty. === Mark Millard markmi at dsl-only.nethelp
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?8B4E345E-1BB0-4D07-9730-740BA48E800F>