- Boost - lists.boost.org

Re: [boost] A fast-pathed rw-mutex algorithm for Boost...
by Dave Handley 26 Oct '06

26 Oct '06

> Date: Tue, 24 Oct 2006 10:48:27 -0700 > From: "Chris Thomasson" <cristom(a)comcast.net> > Subject: Re: [boost] A fast-pathed rw-mutex algorithm for Boost... > To: boost(a)lists.boost.org > Message-ID: <ehliu9$p9$1(a)sea.gmane.org> > > "Dave Handley" <dave.handley(a)morganstanley.com> wrote in message > news:41B9816CBF7F0249AEED3F2B4FE1A5A958B8@lnpdtdc1.pdt.ms.com... > >> Date: Tue, 24 Oct 2006 09:03:13 -0700 > >> From: "Chris Thomasson" <cristom(a)comcast.net> > >> Subject: Re: [boost] A fast-pathed rw-mutex algorithm for Boost... > >> To: boost(a)lists.boost.org > >> Message-ID: <ehlcp0$3vf$1(a)sea.gmane.org> > > [...] > > > I've not looked at the code in question - but the way I see it a well > > behaved readers/writer lock should do the following: > > > > 1) If a single writer is waiting, further attempts to acquire read > > locks should wait until the writer has locked and unlocked. > > 2) If a second writer starts to wait, it should not prevent other read > > locks getting involved - and should only start to block new read locks > > at the point at which it acquires the lock. > > > > If you do this (and have reasonable selection of which waiting thread is > > serviced first) then you can avoid starvation on both sides under all > > circumstances. > > My current code gives writers priority because there is not going to be a > lot of frequent writes anyway... However, it would be trivial to make it > work like you describe... > > > > > > This can easily be done with atomics and a single 32 bit ref count. > > Yeah... However, I don't really have any problems with using DWCAS when I > am > not dealing with pointers... A DWCAS algorithm that does not deal with > pointers on a 32-bit system is 100% compatible with normal CAS on a 64-bit > system... This is fairly portable because nearly all 32-bit architectures > have a DWCAS, but hardly any 64-bit architecture has DWCAS; as long as you > don't use pointers your okay: > > http://groups.google.com/group/comp.arch/browse_frm/thread/71f8e0094e353 e5 > > The way our code operates with a single 32 bit ref count is that the ref count is signed. When a reader wants to acquire the lock, it checks the sign, if 0 or +ve it increments the ref count and acquires the lock. If -ve it waits until the ref count is zero (which happens when the writer releases the lock). When a reader wants to release the lock, it checks the sign, incrementing if -ve and decrementing if +ve. When a writer wants to acquire the lock, it checks the sign. If 0 or +ve, it flips the sign and decrements the ref count (so 0 becomes -1 and 1 becomes -2 etc.). It then waits until the ref count hits -1, at which point it has got the lock. If the sign is -ve it waits. When a writer wants to release the lock, the ref count must be -1, it simply sets it to 0. If we look at the ref count, positive values give us a count of the number of readers that currently hold the lock. Zero means no-one has the lock, and -1 means a single writer has the lock. Other negative values, mean that a writer is waiting, and we have abs(ref_count)-1 readers currently holding the lock preventing the writer from acquiring it. The only information we don't have in this ref count is the numbers of readers or writers waiting on the lock. Dave

2 1

Boost Filesystem and Regex
by Shams 26 Oct '06

26 Oct '06

Hi, Does Boost.Filesystem have "builtin" Boost.Regex support for directory/file iteration, remove, rename etc. or do we have to write wrappers to be able to do something like this? If not then is there an existing implementation anywere? Thanks Shams

2 2

GIL Review
by Christian Henning 25 Oct '06

25 Oct '06

Sorry, I overlooked that I sent my review to the boost user list. I hope it's not too late. Hi Here's my GIL review: > Please always explicitly state in your review, whether you think the > library should be accepted into Boost. I vote to accept GIL into boost. > - What is your evaluation of the design? I like the design and the use of modern C++ paradigms. The design seems to well thought through and taken the whole discourse so far it also seems to hold against heavy criticism. I don't think I'm the position or having the skills to criticize the design. I'm a boost user and happy to have found that place. > - What is your evaluation of the implementation? Seems fine to me. Although I'm a bit anal when code lines that are longer than 80 characters. But that is more minor. I'm not sure the extensions are on review as well. If so, using plain old C code for reading and writing to files, feels outdated. > - What is your evaluation of the documentation? This my biggest problem so far. I like a lot of documentation. My favorite is the one from ASIO. And in this way GIL is loosing. I do understand that documentation is hard but I think it should be treated as source code. For example, I was trying to figure out how to implement indexed images using GIL. Since, it's not there yet it should be easy to add. If I would have known through the documentation the image views have struct add_deref then I might have been able to figure out how to do it. Also I think the documentation should not just enumerate but also explain plus examples. As I said I'm a user and so I don't need to understand all the details on how and why. > - What is your evaluation of the potential usefulness of the library? There are millions of things the lib can do and hopefully will. Image processing is just the start! I'm think of 3D, loading Quake models and so on ... . Would be cool if researchers can use GIL to test new algorithms quickly in an efficient way. > - Did you try to use the library? With what compiler? Did you have > any problems? Yes, I'm using it and have send the creators extensions to read and write bmp files. An early version. Sorry, Lubomir, haven't had a single chance to work on it. I also played around with adding SDL to GIL. Using SDL would make it easy to display a window with the actual image. I'm using VC2003 on Windows XP SP2. > - How much effort did you put into your evaluation? A glance? A quick > reading? In-depth study? I know about GIL for quite a while now. I have tried some things and was usually successful in implementing. But I haven't done an in-depth study. > - Are you knowledgeable about the problem domain? I would be the last to call myself an expert in image processing. I know basics and eager to learn more. The library will certainly help me. Regards, Christian

2 2

Procedure to contriube to existing library
by Guy Peleg 25 Oct '06

25 Oct '06

Hi, Is there a procedure of suggesting contributions to existing libraries? Thanks, Guy __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

5 6

[system] - patch for problems with strerror_r on LINUX/HPUX
by Oliver.Kowalke＠qimonda.com 25 Oct '06

25 Oct '06

Hi, LINUX and HP/UX have different strerror_r functions. At least for me the moduification works. regards, Oliver std::string errno_md( const error_code & ec ) { # if !defined(BOOST_WINDOWS_API) // does not have strerror_t #if defined(__linux) char buf[64]; char * bp = buf; std::size_t sz = sizeof(buf); return strerror_r( ec.value(), bp, sz); #elif defined(__hpux) return strerror( ec.value() ); #else char buf[64]; char * bp = buf; std::size_t sz = sizeof(buf); int result; while ( (result = strerror_r( ec.value(), bp, sz )) == ERANGE ) { if ( sz > sizeof(buf) ) std::free( bp ); sz *= 2; if ( (bp = static_cast<char*>(std::malloc( sz ))) == 0 ) return std::string( "ENOMEM" ); } std::string msg( ( result == EINVAL ) ? "EINVAL" : bp ); if ( sz > sizeof(buf) ) std::free( bp ); return msg; # endif # else const char * c_str = std::strerror( ec.value() ); return std::string( c_str ? c_str : "EINVAL" ); # endif }

2 1

[test] Patch to fix serialisation regressions
by John Maddock 25 Oct '06

25 Oct '06

The following patch to Boost.test clears the serialisation lib failures with VC6 see : http://tinyurl.com/y3qpqf OK to commit (release branch only: I assume cvs HEAD no longer supports VC6 anyway)? John. Index: boost/test/test_tools.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/test/test_tools.hpp,v retrieving revision 1.60.2.1 diff -u -r1.60.2.1 test_tools.hpp --- boost/test/test_tools.hpp 24 Jul 2006 00:43:17 -0000 1.60.2.1 +++ boost/test/test_tools.hpp 25 Oct 2006 18:26:11 -0000 @@ -482,7 +482,7 @@ //____________________________________________________________________________//-struct BOOST_TEST_DECL equal_impl_frwd {+struct /*BOOST_TEST_DECL*/ equal_impl_frwd { template <typename Left, typename Right> inline predicate_result call_impl( Left const& left, Right const& right, mpl::false_ ) const

2 1

Re: [boost] Boost on Windows Mobile
by pdaengr＠hotmail.com 25 Oct '06

25 Oct '06

pdaengr(a)hotmail.com wrote: >>I need to run some of the boost dlls on a PDA (Windows Mobile 5.0). Do I >>need to compile them with a different switch or with a different version >>of bjam or anything like that? >> >>I've compiled the dlls with with VC8 on Windows XP. Would these versions >>of the dlls run on Windows Mobile? >> >We build apps for WinCE 5.0 using VC8 and have successfully used several of >the build-requiring boost libraries including thread (with some minor >fixes), signals, and date_time. We use STLPort 5 to make up for the >incomplete standard C++ library that comes with CE 5. > >-Dave As I'm sure you can tell, I'm fairly new at programming for the PDA with C++. I have been using the .NETCF and C# in the past, but my current project requires the use of C++. **pauses for flames** How did you go about compiling the libraries? Were you able to use a downloaded version of bjam to compile for VC8 + WinCE? Since I am unfamiliar with bjam and all, my first instinct was to put the classes I need into projects in Visual Studio, but at second glance it seemed like quite an undertaking. Then, I found the CreateThread method in windows.h and have been playing with it a bit. I think it will get me through what I need to do with threads, for now, but I'd say I'll need a better implementation soon. Thanks for your help. -Glen _________________________________________________________________ All-in-one security and maintenance for your PC. Get a free 90-day trial! http://clk.atdmt.com/MSN/go/msnnkwlo0050000002msn/direct/01/?href=http://ww…

2 1

www.boost.org
by Jan Stetka 25 Oct '06

25 Oct '06

www.boost.org is still giving an error message, so whatever john maddock did last time, can he do it again. Also we need to find out what un-fixed it!

2 2

Boost regression notification (2006-10-25 [RC_1_34_0])
by Douglas Gregor 25 Oct '06

25 Oct '06

Boost Regression test failures Report time: 2006-10-25T12:33:37Z This report lists all regression test failures on release platforms. Detailed report: http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/i… 47 failures in 9 libraries iostreams (10) lambda (1) mpl (4) parameter (3) program_options (7) random (1) serialization (16) spirit (4) utility (1) |iostreams| bzip2_test: msvc-7.1 msvc-8.0 example_test: vc-6_5-stlport file_descriptor_test: cw-9.4 finite_state_filter_test: cw-9.4 gzip_test: msvc-7.1 msvc-8.0 mapped_file_test: cw-9.4 zlib_test: msvc-7.1 msvc-8.0 |lambda| exception_test: gcc-3_4_4_tru64 |mpl| multiset: gcc-4.0.3_linux gcc-4.1.0_linux gcc-4.1.0_linux_x86_64 gcc-4_0_3_tru64 |parameter| python_test: qcc-3.3.5_gpp sfinae: borland-5_8_2 vc-6_5-stlport |program_options| cmdline_test_dll: cw-9.4 options_description_test_dll: cw-9.4 parsers_test_dll: cw-9.4 positional_options_test_dll: cw-9.4 unicode_test_dll: cw-9.4 variable_map_test_dll: cw-9.4 winmain_dll: cw-9.4 |random| random_test: intel-linux-9.0 |serialization| test_class_info_load_text_warchive: vc-6_5 test_non_default_ctor2_text_archive: qcc-3.3.5_cpp test_non_default_ctor2_text_archive_dll: qcc-3.3.5_cpp test_non_default_ctor2_text_warchive: qcc-3.3.5_cpp test_non_default_ctor2_text_warchive_dll: qcc-3.3.5_cpp test_reset_object_address: vc-7_0 test_reset_object_address_dll: vc-7_0 test_simple_class_binary_archive_dll: vc-6_5 test_simple_class_text_archive_dll: vc-6_5 test_simple_class_text_warchive_dll: vc-6_5 test_simple_class_xml_archive_dll: vc-6_5 test_simple_class_xml_warchive_dll: vc-6_5 test_variant_xml_archive: borland-5_8_2 test_variant_xml_archive_dll: borland-5_8_2 test_variant_xml_warchive: borland-5_8_2 test_variant_xml_warchive_dll: borland-5_8_2 |spirit| action_tests: qcc-3.3.5_cpp qcc-3.3.5_gpp action_tests_debug: qcc-3.3.5_cpp qcc-3.3.5_gpp |utility| operators_test: gcc-3.4.5_linux_x86_64

1 0

[enable_if] Identifying when the GCC workaround is required
by Sylvester-Bradley, Gareth 25 Oct '06

25 Oct '06

http://www.boost.org/libs/utility/enable_if.html#htoc9 explains that some compilers (e.g. GCC 3.2) diagnose two functions as ambiguous if the only distinguishing factor is a different condition in an enabler and offers some workarounds. Is there a BOOST_ macro that's defined when these workarounds need to be applied? If not, any help putting such a macro together would be gratefully received! Thanks, Gareth ************************************************************************ The information contained in this message or any of its attachments may be confidential and is intended for the exclusive use of the addressee(s). Any disclosure, reproduction, distribution or other dissemination or use of this communication is strictly prohibited without the express permission of the sender. The views expressed in this email are those of the individual and not necessarily those of Sony or Sony affiliated companies. Sony email is for business use only. This email and any response may be monitored by Sony to be in compliance with Sony�s global policies and standards

1 0