Date: Mon, 30 Sep 2019 18:03:47 -0700 From: John Kennedy <warlock@phouka.net> To: Brian Bostwick <bostwick.brian@gmail.com> Cc: freebsd-arm@freebsd.org Subject: Re: Building world and kernel on RPI3 B+ compilation failure, unwind-dw2.c. Message-ID: <20191001010347.GB88930@phouka1.phouka.net> In-Reply-To: <20190929164654.GB77308@phouka1.phouka.net> References: <CAM9he-Woj3gkiGPb%2BWTOq7y8kcRjG2RjPBLsXYhk-QayZx75HA@mail.gmail.com> <20190929141211.GA96467@phouka1.phouka.net> <CAM9he-XnPTbK_aGPzWbaZx_N6%2B-x9CUck4Nz0zvuOSYgukLX0A@mail.gmail.com> <20190929164654.GB77308@phouka1.phouka.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Sep 29, 2019 at 09:46:54AM -0700, John Kennedy wrote:
> Well, It's a RPI3B+, so I'll let you know in ~28-31 hours, give or take what
> point in the compile that fragment lands in.
I had no problems crunching BETA2 from BETA1.
FreeBSD 12.1-BETA2 #41 r352780+b78e5b46c44c(releng/12.1)-dirty: Mon Sep 30 11:10:03 PDT 2019
What does your src.conf look like? WITHOUT_LLVM_LIBUNWIND set? Maybe
anti-LLVM/pro-GCC in some other way?
From src.conf(5):
WITHOUT_LLVM_LIBUNWIND
Set to use GCC's stack unwinder (instead of LLVM's libunwind).
This is a default setting on arm/arm, arm/armv6, arm/armv7,
powerpc/powerpc, powerpc/powerpc64, powerpc/powerpcspe and
sparc64/sparc64.
You had problems in /usr/src/gnu/lib/libgcc, which seems to be triggered on
something like that (I really haven't tried to delve that deep into the FreeBSD
source Makefiles). Maybe copied configuration settings over from a previous
RPI? Your RPI3B+ is arm64, so if you had that set you might want to yank it.
[/usr/src/gnu/lib/Makefile]
.if ${MK_LLVM_LIBUNWIND} == "no"
SUBDIR+= libgcc
.endif
My build didn't go down into that directory at all.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20191001010347.GB88930>