- Boost-users - lists.boost.org

boost::thread not started
by justin_michel 11 Mar '03

11 Mar '03

Platform : Windows XP, Pentium 4, Visual C++ 6 sp5 I've traced down my problem with starting a thread from a DLL called from Java JNI. If you look at the source code thread.cpp in boost, you can see that a condition variable is used block the creating thread until the new thread function is started. Basically, what I am seeing is that if you wait on this condition variable, then the thread will never start. If I remove the call to param.wait() at the bottom of the constructor then everything works fine. I can't think of a logical explanation, but it's very repeatable. I was also able to recreate this problem using the latest source from CVS. The simple solution would be to remove the condition variable from thread_param code entirely. This would simplify everything, and I don't really see the value anyway. All that is currently guaranteed is that thread_proxy will be called before the boost::thread() constructor is finished. But it doesn't guarantee that threadfunc() will be called before the constructor is finished. Thanks, Justin Michel OCI

2 3

uBLAS: Can I make my vector<...> derived class 'copy constructible' ?
by speedsnaii 11 Mar '03

11 Mar '03

Hi Joerg, or anyone who can help! As I am still on my spline templates, I came on to the following beast: template<class T> struct index_vector : public vector<std::pair<T, vector<ublas::vector<T>::size_type> > > {}; This should be an index vector that relates to a double (first) a vector of indices (second) which happens to be different for each 'first'. So far so easy. I can say: index_vector<double> IY; IY.resize(5); IY[0].first = 3.14; IY[0].second.reset(vector<double>(3)); IY[0].second[2] = 1; But I cannot: index_vector<double> IZ(IY); or index_vector<double> IZ; IZ = IY; or IZ.reset(IY); This compiles, but gives me runtime assertions because of size mismatches. I think my derived class is not 'copy constructible' and 'assignable'. Is there a simple way to fix this? (I am not even sure whether this is a uBLAS matter. So please simple tell me, and I will try to learn the missing parts.) Thank you, Roland

2 2

How to get extents of multi_array dimensions
by Neil Mackie 11 Mar '03

11 Mar '03

I am using multi-array for the first time, without problems. Now I want to pass the multi_array around and the users of it need to find out what the maximum array indexes are in each dimension. But I do not know how to obtain them from the object. I have read the documentation and the header file but if there are any clues there I have missed them. I have also searched the archives of this list for relevant posts. Is there a call on multi_array to return the array bounds and if so how do you specify which dimension you are requesting the bounds of? Does anyone have an example they could post? Regards, Neil (J Mackie).

1 0

index(*vi) in the BGL quick tour?
by Jack Rabbit 11 Mar '03

11 Mar '03

Hi, My IDE is VC++6.0. I'm on win2k, using boost 1.29.0. I'm trying to get the quick tour (BOOST_HOME/libs/graph/doc/quick_tour.html) of the graph library to compile, but am experiencing a problem. In the example on how to invoke Dijkstra's algorithm, it seems we're calling a function index(*vi). I'm guessing it's supposed to get the node number for the node at vertex_iterator vi. The problem is that my compiler can't find the function ('index' : undeclared identifier). Doing a manual 'search in files' doesn't reveal any legible candidates either. Is this an error in the docs or am I missing something bigger here? kind regards, Dieter Vrancken

1 0

spirit framework...
by Michael Burbidge 10 Mar '03

10 Mar '03

I understand that the spirit compiler framework has been accepted as part of boost. Does anyone know when it will be part of the standard boost distribution. I'm finding that spirit 1.5.1 generally works on Mac OS X using CodeWarrior, but that I'm experiencing runtime assertions running some of the examples. In particular the ast_xml example. Thanks, Michael-

2 1

Problem with tuples::ignore
by Tom Matelich 10 Mar '03

10 Mar '03

I'm sure I've seen posts about this, but have been unable to track it down... Apparently gcc 3.2 only uses the cpp file name to "anonymize" anonymous namespaces. My problem is that Application A links libs B and C, which both have files named foo.cpp. These foo.cpp's both include boost/tuples/details/tuple_basic.hpp, thus causing duplicate symbols of tuples::{anonymous}::ignore. Obviously I can work around this by renaming the cpp files, but I thought I'd see if there was a quicker solution. Thanks Tom pardon the disclaimer... ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. -----------------------------------------------------------------------

1 0

uBLAS: reference to container data & iteratordeclaration
by speedsnaii <speedsnaii＠yahoo.de> 08 Mar '03

08 Mar '03

Hi, *) First question: I am trying to get a reference to data of a container in the following way: matrix<double> m(3,3); vector<double>& v(m.column(2)); But this seems not to be possible. Is there a way to achieve this behaviour? What I basically want to achive is having functor object that holds a reference to a row/column of my container matrix. >From my questions about argument types I know that m.column(2) is a kind of an expression. How can I get the underlying reference to data if there even is one? *) Second question: matrix<double> m(10,3); ... fill the matrix ... matrix<double>::iterator it; // wrong this does not work!! for (it = m.column(2).begin(); it != m.colum(2).end(); ++it) { cout << *it << " " << foo(*it) << endl; } How would I declare my iterator? I understand when I use std::library-algorithms the automatic template parameter defer will give me the correct type from begin(). But in my case how should I declare the iterator? Thank you for help. Roland

3 10

thread problems with Java JNI
by justin_michel 08 Mar '03

08 Mar '03

A simple test program using boost::thread works fine, but when I try to use boost::thread in a DLL library which is called from a Java thread, the thread blocks forever in the constructor for boost::thread. Looking at the source code for boost::thread, I see that it blocks on a condition waiting for the thread to start. Maybe the thread never starts? So. I modified my code to use _beginthreadex() directly to start my worker thread. I hoped to at least get a better idea of what was going wrong, but the code worked flawlessly. Are there known problems with using boost::thread from a DLL? Any other ideas? Thanks, Justin Michel

1 0

boost-build path discrepancies
by Michael Hunley 08 Mar '03

08 Mar '03

As per my recent post "Installing boost libs for dependant apps," I recently used the boost-build system for my own apps make setup. I found it to be very powerful and useful (more than I could use to begin with). I used the version 1 build system becuase it looked like it would be easier to get up and running, but plan to migrate to v2 soon since it is more powerful and I much prefer the dependency handling of adding sub-projects by name, not path to a Jamfile. But I need to learn it first. I did this on RedHat 8.0 using gcc 3.2.2 with a cut of the Boost CVS from 2/6/2003 on the Boost-Build v1.30.0. I copied most of the main files (features.jam, boost-build.jam, etc) from the main boost build setup in BOOST_ROOT and BOOST_ROOT/tools/build. I ran into two issues with pathing and another with the LinkLibraries rule. I tried a couple of approaches to including externally built libraries to my build hierarchy (several sub-project libs and one app). When I tried LinkLibraries it didn't complain, but the library never got added to my link action in the NEEDLIBS variable. I tried tracing it down and added Echo actions to the LinkLibraries to see what it was adding, which was nothing, but I could have mis-understood the NEEDLIBS on $(<) expression. Regardless, they did not appear in the list of libs passed to the linker. I then tried the <libraryfile> feature in my requirements section. That worked fine. So, is LinkLibraries deprecated (the docs did not mention it that I could find)? I had it first in my list, before my exe target, but in the same file. Was it supposed to go after? I liked LinkLibraries better than the <libraryfile> since it would do the name mangling for me (prefix="lib", suffix=".so" under gcc). The other problem I ran into revolved around use of relative paths and I believe my results are counter to what the documentation claims. First, when I added dependant libraries as source entries in my exe target the docs claim I can make those paths absolute. My setup is a little out of the norm in that my app is at the leaf node of the directory tree, so the relative path to the main lib it needed was $(DOTDOT). However, I have several libs common to other apps, but in the same boost build hierarchy, so those are relative to $(BOOST_ROOT). When I tried having a source line: <lib>$(BOOST_ROOT)/Common/CommonLib from within my Jamfile located at $(BOOST_ROOT)/ServerLib/Server The output when I tried bjam was: (|ServerLib|Server|home|michael|Proj|Common) not found (I am only sure of the folder part of that, the exact error syntax may be slightly off). I had to specify the paths as: $(DOTDOT)/$(DOTDOT)/Common/CommonLib In the second case, it was relative pathing that wouldn't work. I specified most of the common features and settings using template targets that I then imported into each Jamfile (Very nice feature). The one of concern was located in the $(BOOST_ROOT)/ServerLib folder. In the $(BOOST_ROOT)/ServerLib/Server/Jamfile I added a requirement: <include>$(DOTDOT) To get back to ServerLib which it depends on. I swear that worked the first time I got it built, but last night when I did a rebuild all (bjam -a) it failed and claimed it could not find include files. I ended up having to change it to absolute pathing to work. Are the docs wrong on the two pathing issues? Is there something I could be setting up wrong to get the behavior I saw? thanks. michael

1 0

installing boost libs for dependant apps
by Michael Hunley 08 Mar '03

08 Mar '03

I just completed porting our server from XP to Linux, which uses Boost::Thread for the threading library. I found it was infinitely more convenient to publish the libs in their various formats (variant|sub-variant) into a common, top-level folder under BOOST_ROOT. This made it much more managable to both link it into my application (without having to grist the folder name) and add the proper path to my LD_LIBRARY_PATH env var so it could find the shared libraries when I ran. I don't know if there is a prescribed way of doing this already in place, but I did not see anything in the build documentation (like bjam "Install [folder]" or some such). I ended up modifying the BOOST_ROOT/Jamfile to have a stage target that published all the threading libs to BOOST_ROOT/bin with proper suffixes for the variants I cared about (debug and release for the moment). I still had to append the proper suffixes in my own Jamfile (I used BJam for my make as well), but it was a heck of a lot better than a 10 deep folder hierarchy. I have a couple of question and two requests. For reference I am using a cut from CVS around 2/6/2003. How are others doing this sort of thing that are using Boost? Is there a common way to do what I did by doctoring the Jamfile that is less invasive (so I don't have to repeat in the future when I get releases from CVS)? I am new to Linux & Boost, so may not know how I should be getting at these things. My request, if the answers to the above don't address them, is that some form of install or publish to a common location be added to the boost build. If I were more intimately familiar with Boost and the build system (and the various installation platforms) I would volunteer to help, but... My second request may also already be possible, but my limited experiment and reading made it seem not. I'd like to be able to declare a sub-project that points to a whole other boost-build.jam hierarchy; e.g. I want to make Boost a sub-project of mine. This seems much more plausible and likely in the new versoin 2 format where the dependency declarations are more straight-forward (by name, not explicit path). I await your feedback. Thanks. michael

1 0