Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 27 Apr 2019 16:55:08 -0700
From:      bob prohaska <fbsd@www.zefox.net>
To:        Jan Beich <jbeich@FreeBSD.org>
Cc:        freebsd-ports@freebsd.org
Subject:   Re: stack protector mode differs in PCH file vs. current file:
Message-ID:  <20190427235508.GA86998@www.zefox.net>
In-Reply-To: <wojf-tw95-wny@FreeBSD.org>
References:  <20190426232058.GA80256@www.zefox.net> <wojf-tw95-wny@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Apr 27, 2019 at 03:26:14PM +0200, Jan Beich wrote:
> bob prohaska <fbsd@www.zefox.net> writes:
> 
> > In trying (still) to compile www/chromium on an RPI3 running -current with
> > ports at 500082 and system at 346613 portmater is stopping in (I think)
> > openjdk8 with
> > errorerrorerror: : errorstack protector mode differs in PCH file vs. current file: : 
> > stack protector mode differs in PCH file vs. current filestack protector mode differs in PCH file vs. current file
> > stack protector mode differs in PCH file vs. current file
> 
> Can't say much without full build log but it maybe a regression from
> https://svnweb.freebsd.org/changeset/ports/499897
> 
> Maybe precompiled.hpp.pch is generated with different arguments compared
> to when it's included in source files. Try the following workaround:
> 
> --- java/openjdk8/Makefile.orig
> +++ java/openjdk8/Makefile
> @@ -203,14 +203,14 @@ CONFIGURE_ENV+=	LIBCXX="-lc++"
>  .endif
>  .endif
>  
> -# GCC is broken with PCH: https://lists.freebsd.org/pipermail/svn-src-all/2015-March/101722.html
>  .if ${COMPILER_TYPE} == gcc
>  CONFIGURE_ARGS+=	--with-toolchain-type=gcc
> -.if ${ARCH} == "powerpc64"
> -MAKE_ARGS+=	USE_PRECOMPILED_HEADER=1
> -.else
> -MAKE_ARGS+=	USE_PRECOMPILED_HEADER=0
>  .endif
> +
> +# GCC is broken with PCH: https://lists.freebsd.org/pipermail/svn-src-all/2015-March/101722.html
> +# PCH is poorly tested outside of x86
> +.if ${ARCH} != "amd64" || ${COMPILER_TYPE} == gcc
> +MAKE_ARGS+=	USE_PRECOMPILED_HEADER=0
>  .endif
>  
>  .if empty(ICONV_LIB)

I'm doing something wrong, patch replies
patch: **** malformed patch at line 22:  .if empty(ICONV_LIB)
and exits without doing anything. Come to think of it, shouldn't
.if [anything] 
be followed by
.endif
eventually?

Thanks for your help!

bob prohaska
 



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20190427235508.GA86998>