|
Boost-Build : |
Subject: [Boost-build] [build] [context] compile failures (was: [1.51.0] Beta 1 release candidates available)
From: Eric Niebler (eric_at_[hidden])
Date: 2012-08-01 10:33:57
(cc'ing boost-build)
On 8/1/2012 5:18 AM, Pekka Seppänen wrote:
> On 1.8.2012 5:40, Marshall Clow wrote:
>> Release candidate files for 1.51.0 beta 1 are available at
>> http://boost.cowic.de/rc/
>>
>> As always, the release managers would appreciate it if you download
>> the candidate of your choice and give building it a try. Please report
>> both success and failure, and anything else that is noteworthy.
>>
>> This helps ensure the candidates build OK before we push them out to
>> SourceForge.
>
> Hello,
>
> at least MSVC (could be others, too, this is a compiler specific jam-file
> issue) fails to fully compile Boost.Context if there's any user specified
> preprocessor define passed to B2 that has characters that command prompt may
> interpter as control characters, e.g.
> ``define=BOOST_USER_CONFIG="<SuperiorConfig>"'' that is.
>
> The problem lies in how tools\build\v2\tools\msvc.jam handles .ASM (and .RC
> files, too, not that Boost yet uses any during the build) files:
>
> line 314:
> actions compile.asm
> {
> $(.ASM) -c -Zp4 -Cp -Cx -D$(DEFINES) $(ASMFLAGS) $(USER_ASMFLAGS) -Fo
> "$(<:W)" "$(>:W)"
> }
>
> As command line is not passed thru a .RSP file (just a normal text file where
> the rest of the command line arguments are read) unlike compile.asm's (and
> .rc's) counterparts do, characters such as "<" and ">" affect the control
> flow, which in this case is not what user wants (at least not what *I* want :D).
>
> So, as a fix I propose that either do a full .RSP treatment or simply enclose
> defines inside a double quotes ("-D$(DEFINES)"); That should fix it.
>
> (Not sure if $(INCLUDES) should be passed to $(.ASM), in this particular
> example it's the not the root cause anyway [<SuperiorConfig> not being found
> due to missing include path] as no .ASM file currently actually includes any
> [external] file.)
Thanks for the report Pekka. This doesn't seem like something that
should hold up the beta, but hopefully someone can fix it for release.
-- Eric Niebler BoostPro Computing http://www.boostpro.com
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk