Hi,
> If you want to dip a toe into the compiler (since most of us are
> struggling to get the release out this week), a simple fix might be to
> have the _mt identifiers that the compiler generates increment a global
> counter to avoid them colliding with one antoher. I'm not entirely sure
> why nesting classes makes this happen more than sibling classes at module
> scope would be. Grepping on "_mt" in the compiler makes it pretty easy to
> find the dozen or so places where they're created (which could/should
> arguably be refactored into a utility routine defining the global
> counter).
I gave this a try, but can't get it work. Well, this definitely isn't an urgent
for me so I don't need fix right now...
I wrote a helper method makeArgSymbol, which returns new ArgSymbol.
makeArgSymbol has a counter, so whenever iName argument is "_mt", it appends an
unique number to the name. It also does that whenever iName is "outer", since
that was the next error I got after fixing the _mt case. Now I do get the error
below.
And as I said, this isn't urgent, so feel free to ignore this until you are
done with the new release.
machine info: Linux alks-main 3.13.0-45-generic #74-Ubuntu SMP Tue Jan 13
19:36:28 UTC 2015 x86_64
CHPL_HOME: /home/akseli/projects/chapel/chapel
script location: /home/akseli/projects/chapel/chapel/util
CHPL_HOST_PLATFORM: linux64
CHPL_HOST_COMPILER: gnu
CHPL_TARGET_PLATFORM: linux64
CHPL_TARGET_COMPILER: gnu
CHPL_TARGET_ARCH: unknown
CHPL_LOCALE_MODEL: flat
CHPL_COMM: gasnet
CHPL_COMM_SUBSTRATE: udp
CHPL_GASNET_SEGMENT: everything
CHPL_TASKS: fifo
CHPL_LAUNCHER: amudprun
CHPL_TIMERS: generic
CHPL_MEM: cstdlib
CHPL_MAKE: make
CHPL_ATOMICS: intrinsics
CHPL_NETWORK_ATOMICS: none
CHPL_GMP: gmp
CHPL_HWLOC: none
CHPL_REGEXP: re2
CHPL_WIDE_POINTERS: struct
CHPL_LLVM: none
CHPL_AUX_FILESYS: none
init : 0.044 seconds [ 450]
parse : 0.288 seconds [ 179700]
checkParsed : 0.121 seconds [ 179700]
docs : 0.127 seconds [ 179700]
readExternC : 0.123 seconds [ 179700]
expandExternArrayCalls : 0.108 seconds [ 179700]
cleanup : 0.145 seconds [ 179960]
scopeResolve : 0.867 seconds [ 224969]
flattenClasses : 0.165 seconds [ 224969]
normalize : 0.778 seconds [ 435951]
checkNormalized : 0.371 seconds [ 435951]
buildDefaultFunctions : 0.614 seconds [ 473934]
createTaskFunctions : 0.430 seconds [ 477212]
$CHPL_HOME/modules/internal/ChapelIteratorSupport.chpl:119: error: unresolved
call 'AbstractRootLocale.initOnLocales(tag=iterKind)'
[functionResolution.cpp:2464]
$CHPL_HOME/modules/internal/ChapelIteratorSupport.chpl:119: note:
instantiated from _toLeader [functionResolution.cpp:2471]
$CHPL_HOME/modules/internal/ChapelIteratorSupport.chpl:123: note:
instantiated from _toLeader [functionResolution.cpp:2471]
$CHPL_HOME/modules/internal/ChapelLocale.chpl:211: note: candidates are:
AbstractRootLocale.initOnLocales() [68252] [functionResolution.cpp:2479]
$CHPL_HOME/modules/internal/ChapelLocale.chpl:224: note:
AbstractRootLocale.initOnLocales(param tag: iterKind) [68331]
[functionResolution.cpp:2479]
$CHPL_HOME/modules/internal/ChapelLocale.chpl:241: note:
AbstractRootLocale.initOnLocales(param tag: iterKind, followThis) [68484]
[functionResolution.cpp:2479]
------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Chapel-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/chapel-users
