- Boost - lists.boost.org

shared_ptr.hpp & Intel C++ error
by Ralf W. Grosse-Kunstleve 02 Jun '08

02 Jun '08

I'm routinely testing Boost.Python compilation with Intel C++ under Fedora 3 (64-bit). Today I tried Fedora 5, 6, 9, and I'm getting the error below. It can be reproduced by simply including boost/shared_ptr.hpp. I tried adding -pthread but it didn't make a difference. Is this a known problem? Is there a trick to get around it? Ralf P.S.: Intel C++ 9.1 behaves the same way. boost trunk svn revision 46049 % cat include_boost_shared.cpp #include <boost/shared_ptr.hpp> % icpc -V Intel(R) C++ Compiler for applications running on Intel(R) 64, Version 10.1 Build 20080312 Package ID: l_cc_p_10.1.015 Copyright (C) 1985-2008 Intel Corporation. All rights reserved. FOR NON-COMMERCIAL USE ONLY % icpc -I/net/rosie/scratch1/rwgk/hot/boost -c include_boost_shared.cpp /net/rosie/scratch1/rwgk/hot/boost/boost/detail/spinlock_sync.hpp(36): error: identifier "__sync_lock_test_and_set" is undefined int r = __sync_lock_test_and_set( &v_, 1 ); ^ /net/rosie/scratch1/rwgk/hot/boost/boost/detail/spinlock_sync.hpp(50): error: identifier "__sync_lock_release" is undefined __sync_lock_release( &v_ ); ^ compilation aborted for include_boost_shared.cpp (code 2)

3 2

Re: [boost] shared_ptr.hpp & Intel C++ error
by Ralf W. Grosse-Kunstleve 02 Jun '08

02 Jun '08

Thanks for the reply! Unfortunately, it doesn't help in my case, because I have a x86_64 platform. I blindly tried all includes I could find in the Intel C++ directory, but none helped. A look at the docs also didn't lead to any insights... Is there something else I could try? > Have you considered running the smart_ptr and bind tests on a regular basis > and uploading the results ;-) Yes, I could do that. Ralf

2 1

Re: [boost] Rounding, Truncating Values
by Tom Brinkman 02 Jun '08

02 Jun '08

>> I'm not even sure I understand the contract. How can you >> round to a number of decimal places when the result is not >> necessarily representable? Good question. Not sure I have an answer. I just evolved these algorithms becouse I couldnt find them anywhere else. >> Rather than an implementation, could you >> provide an example of a documentation, stating precisely what the >> return value should be? Sure. I could do that. >> Also, if the intention is float/double/long >> double, why templates, as opposed to three overloads? Thanks for any >> clarification. Well, this is boost after all. What tradeoffs do you see? >> Equally useful as rounding, perhaps, would be printing to a string >> with guaranteed rounding. You know that printing or reading a binary >> floating point to a decimal string representation, and back, may not >> result in the same number. This goes with printf and scanf, as well >> as I/O streams (perhaps this has been corrected but I haven't seen >> anything very sophisticated in the STL implementations I lurk in). >> There are known algorithms to do this with guaranteed rounding, e.g.: I'm not an expert here. If you would be willing to advise me on these issues, I might be persuaded to put together some test cases and example files. I know there alot of subtle issues involved with rounding floating point numbers.

1 0

[optional] References
by Roman Odaisky 01 Jun '08

01 Jun '08

Greetings, boost::optional<T &> seems to be implemented in terms of a boolean flag and an aligned_storage. Wouldn’t it make more sense to make boost::optional<T &> a thin wrapper over T*? This would save a couple of bytes, speed up get_ptr by a couple of ticks, and perhaps reduce compilation times. Also, from the documentation[1]: > Rebinding semantics for the assignment of initialized optional references has been chosen to provide consistency among initialization states even at the expense of lack of consistency with the semantics of bare C++ references. It is true that optional<U> strives to behave as much as possible as U does whenever it is initialized; but in the case when U is T&, doing so would result in inconsistent behavior w.r.t to the lvalue initialization state. optional<U> does not have the semantics of U, but rather those of U*. That explains the assignment semantics perfectly well: int a = 1; int b = 2; int& ra = a; int& rb = b; int* pa = &a; int* pb = &b; boost::optional<int &> oa(ra); boost::optional<int &> ob(rb); ra = rb; // assigns b to a pa = pb; // rebinds pa to b oa = ob; // rebinds oa to b -- [1] http://tinyurl.com/55s8f7 -- WBR Roman.

1 0

[Graph] compile error while inheriting boost graph
by Sandeep Gupta 01 Jun '08

01 Jun '08

Hi, My attempt to inherit boost graph failed. The compiler complains that it is not aware of boost graph classes. Would appreciate if someone can tell if this due to boost intricacies or just regular mistake in the syntax. //Declare boost graph within namespace boostd namespace boostd { typedef adjacency_list<listS, listS, ... > Graph; typedef graph_traits<Graph>::vertex_descriptor Vertex; } class DerivedGraph: public boostd::Graph{ public: class Node: public boostd::Vertex{ //error on this line }; }; The compiler complains error: expected class-name before '{' token Thanks sandeep

2 1

Non pre-emptible Boost.Thread
by Edd Dawson 31 May '08

31 May '08

Hello, A while ago, I implemented the Boost.Thread API in terms of Windows Fibers and <uncontext.h> functionality. The result is the ability to run code written with boost threads in a single OS thread. I find it useful for writing tests where determinism is important i.e. to test algorithmic correctness in isolation from synchronization- related concerns. I haven't updated it to mirror the new 1.35 API yet, but I was wondering if any other boost users would be interested in this or if it might be thought suitable for inclusion in the boost library collection. More information is available here: http://www.mr-edd.co.uk/?page_id=91 Direct link to source code: http://www.mr-edd.co.uk/files/coco/coco-0.2.0.zip Kind regards, Edd

5 13

[thread] warnings when linking with the dynamic thread library on cygwin
by vicente.botet 31 May '08

31 May '08

Hello, I have a warning when linking with the thread library. /usr/lib/gcc/i686-pc-cygwin/3.4.4/../../../../i686-pc-cygwin/bin/ld: warning: auto-importing has been activated without --enable-auto-import specified on the command line. When I set <library>/boost/thread//boost_thread/<link>static; instead of <library>/boost/thread//boost_thread; this works without warnings. What I need to do to link dynamically without warnings? Best regards Vicente ================== Command $ bjam The jamfile contains using testing ; if ! $(BOOST_ROOT) { BOOST_ROOT = [ modules.peek : BOOST_ROOT ] ; } project : : requirements <include>$(BOOST_ROOT) <include>../../.. <threading>multi <target-os>cygwin <threadapi>pthread <variant>debug <library>/boost/test//boost_unit_test_framework/<link>static <threading>multi <library>/boost/thread//boost_thread; Link warning: gcc.compile.c++ ..\..\..\bin.v2\libs\sync\test\include_concurrent_component.test\gcc-3.4.4\debug\target-os-cygwin\threadapi-pthread\threading-multi\include_concurrent_component.o **passed** ..\..\..\bin.v2\libs\sync\test\include_concurrent_component.test\gcc-3.4.4\debug\target-os-cygwin\threadapi-pthread\threading-multi\include_concurrent_component.test gcc.compile.c++ ..\..\..\bin.v2\libs\sync\test\static_thread_group_test.test\gcc-3.4.4\debug\target-os-cygwin\threadapi-pthread\threading-multi\static_thread_group_test.o gcc.link ..\..\..\bin.v2\libs\sync\test\static_thread_group_test.test\gcc-3.4.4\debug\target-os-cygwin\threadapi-pthread\threading-multi\static_thread_group_test.exe /usr/lib/gcc/i686-pc-cygwin/3.4.4/../../../../i686-pc-cygwin/bin/ld: warning: auto-importing has been activated without --enable-auto-import specified on the command line. This should work unless it involves constant data structures referencing symbols from auto-imported DLLs.Info: resolving vtable for boost::detail::thread_data_baseby linking to __imp___ZTVN5boost6detail16thread_data_baseE (auto-import) testing.capture-output ..\..\..\bin.v2\libs\sync\test\static_thread_group_test.test\gcc-3.4.4\debug\target-os-cygwin\threadapi-pthread\threading-multi\static_thread_group_test.run 1 fichier(s) copi?(s). **passed** ..\..\..\bin.v2\libs\sync\test\static_thread_group_test.test\gcc-3.4.4\debug\target-os-cygwin\threadapi-pthread\threading-multi\static_thread_group_test.test ...updated 6 targets... --------------------------- Vicente Juan Botet Escriba

1 0

[thread] is there an interestoin a static thread group?
by vicente.botet 31 May '08

31 May '08

Hi, The Boost.Thread library provides a thread_group class for a collection of threads that are related in some fashion. This collection is dynamic, i.e. we can add and remove threads to this collection. When this collection is known at the group creation and can not change a static thread group could be useful. This template class thread_tuple_n will have the same operations than a thread_group, join_all, interrupt_all and size and could be used as follows: void free_function() { // .... } class X { public: void mf() { // .... } }; class Y { public: void operator()() { // .... } }; X x; Y y; thread_tuple_3 tt(free_function, make_pair(&X::mf, x), y); // ... tt.interrupt_all(); // ... tt.join_all(); Best _____________________ Vicente Juan Botet Escriba

1 1

[announce] Multi signature boost::function v1.0
by Marco Costalba 31 May '08

31 May '08

Hi all, this is the first official version of multi-signature boost::function extension (MSF) pushed to Boost Vault under the name msf-1.0.zip A multi-signature boost:function 'f' allows something like this int foo1(char) { return -1; } char foo2(string) { return 'x'; } f = foo1; f = foo2; assert( f('x') == -1 ); assert( f("hello") == 'x' ); New in this version: - Feature complete: all boost::function API has been naturally extended - Documented and added a tutorial according to boost standards, see tutorial.html - Added a dynamic dispatcher and an object factory examples that use MSF as a building block - Rewritten part of the code to be even simpler and easy to understand - Fixed case of functors with const and/or volatile operator() - Compiler friendly: tested with gcc3.4, gcc4.2, MSVC - No known issues This library extension, although flexible and IMHO powerful is notably very simple and small, so I would like to ask if someone is interested to add this code in some way directly to boost.function Thanks Marco

7 48

[filesystem] "leaf"
by David Abrahams 31 May '08

31 May '08

I was just reviewing the filesystem docs and came across "leaf()". I'm sure this isn't the first time I've seen it, but this time I picked up a little semantic dissonance. Normally we think of "leaf" in the context of a tree as being a thing with no children. An interior node like a directory that has files or other directories in it is usually not called a "leaf." I wonder if this is the best possible name? Is there a precedent we can draw on in some other language/library? In python, it's os.path.basename(p). Perl, php, and the posix basename command seem to do something similar. -- Dave Abrahams BoostPro Computing http://www.boostpro.com

16 58