March 2006 - Darwin-dev - lists.apple.com

zones in NKEs
by Michael Tuexen 25 Mar '06

25 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Dear all, Best regards Michael _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I'm using zones (allocated with zinit) in an NKE. How can I deallocate them when the NKE is unloaded? This email sent to site_archiver(a)lists.apple.com

1 0

AuthServer StartupItem & File Paths
by Stéphane Sudre 24 Mar '06

24 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Stupid question: Is it normal that the script for the AuthServer Startup Item is calling: tim instead of /usr/sbin/tim ? _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… It seems strange to use a relative file path for the authserver module when other modules are using absolute file path. This email sent to site_archiver(a)lists.apple.com

1 0

Serious problem with gcc 4 on Mac OS X, when using boost:shared_ptr
by Ananda Tallur 24 Mar '06

24 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello, Below is the sample test program : --- macosx_gcc4problem.cpp --- #include <iostream> #include <boost/shared_ptr.hpp> using namespace std; using namespace boost; struct A { int value; ~A () { std::cout << "A destructor" << std::endl << std::flush; } }; struct B { B (shared_ptr <A> val) : a (val) {} shared_ptr <A> a; }; struct C { C (B val) : b (val) {} B b; }; struct D { D (C val) : c (val) {} C c; }; D create () { boost::shared_ptr <A> a (new A); a->value = 12345; return D (C (B (a))); } Output : When run, the value displayed is 13691, instead of 12345 as it should. Thanks Anand _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… Originally, I had discovered a problem when using boost::unit_test_framework, for which I have already posted. But, by further analysing the problem, I could come out to a simple program, using only boost shared pointers, and which fails : -> on Mac OS X 10.4 -> when compiled with gcc 4.0.1 (coming with XCode 2.2.1), or gcc 4.1 (delivered by darwin ports) -> with optimisations : at least -O, or -O1, or -O2,etc... It works correctly with -O0 int main (int argc, char * argv []) { D d = create (); std::cout << "value = " << d.c.b.a->value << std::endl << std::flush; return 0; } --- macosx_gcc4problem.cpp --- To compile it : g++-4.0 -O2 -I <path-to-boost-includes> macosx_gcc4problem.cpp -o macosx_gcc4problem Analysis : -> there is a problem in shared_ptr reference counting, in line : return D (C (B (a))); Just after D creation, the reference counter is at 1, instead of 2 as it should. And when leaving create () function, the automatic variable a is detroyed, which changes the reference counter to 0 and detroys the A object. In main function after call to create (), the reference counter should be of 1, and the object A should be available on the heap. But it has already been destroyed, leading to an incorrect result being displayed. In worst cases, a segmentation fault may happen. Remark : if changing the constructor of B, so that a const reference to the shared_ptr is passed, instead of the shared_ptr itself, it works : --- struct B { B (const shared_ptr <A> &val) : a (val) {} --- To further investigate the problem, it may be needed to analyse the assembly code generated by the gcc 4 compiler. The problem is clearly in the line : return D (C (B (a))); There may be a problem with management of temporary variables, combined with optimisations. I require the help of other developpers, because I think this problem could concern many of us, and not only boost users. This email sent to site_archiver(a)lists.apple.com

1 0

Re: boost test_unit problem when using gcc 4.0.1
by Ananda Tallur 24 Mar '06

24 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello, Below the new test program : --- macosx_gcc4problem.cpp --- #include <iostream> #include <boost/shared_ptr.hpp> using namespace std; using namespace boost; struct A { int value; ~A () { std::cout << "SampleTest destructor" << std::endl << std::flush; } }; struct B { B (shared_ptr <A> val) : a (val) {} shared_ptr <A> a; }; struct C { C (B val) : b (val) {} B b; }; struct D { D (C val) : c (val) {} C c; }; D create () { boost::shared_ptr <A> a (new A); a->value = 12345; return D (C (B (a))); } Output : When run, the value displayed is 13691, instead of 12345. Thanks Anand On 22 févr. 06, at 17:32, Ananda Tallur wrote: Hello, I would like to submit a problem we recently discovered, when using : The problem does not occur when : * using Apple gcc 3.3, both with or without optimization * not using optimization (-O0) Below is the sample program : ---- file boost_unit_test_gcc4_pb_sample.cpp --- #include <iostream> #include <boost/test/unit_test.hpp> using boost::unit_test_framework::test_suite; using boost::unit_test_framework::test_case; using namespace std; class SampleTest { public: SampleTest (int val) : a (val) { } int a; void sampleTest () { std::cout << "a : " << a << std::endl; BOOST_CHECK_EQUAL (a, 12345); } }; When compiled when the above command, and running the program : export DYLD_LIBRARY_PATH=${BOOST_PATH}/lib ./boost_unit_test_gcc4_pb_sample *** 1 failure detected in test suite "Sample test" --- When compiled without optimization (-O0), and running, the result is : ---- Running 1 test case...a : 12345 *** No errors detected ---- Any help would be greatly appreciated ! Regards, Anand _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I have further investigated on this issue, by analysing what happens in the unit_test_framework. I could come out with a simple test program, only using shared pointer, which fails on Mac OS X, when compiled with gcc 4.0.1 (coming with latest XCode 2.2.1), or gcc 4.1 from darwinports, when optimization is enabled (-O, -O1, or -O2). I am using boost 1.33.1, but is it the same with 1.33.0 and 1.32.0. int main (int argc, char * argv []) { D d = create (); std::cout << "value = " << d.c.b.a->value << std::endl << std::flush; return 0; } --- macosx_gcc4problem.cpp --- To compile it : g++-4.0 -O2 -I <path-to-boost-includes> macosx_gcc4problem.cpp -o macosx_gcc4problem Analysis : -> there is a problem in shared_ptr reference counting, in line : return D (C (B (a))); Just after D creation, the reference counter is a 1, instead of 2 as it should. And when leaving create () function, the automatic variable a is detroyed, which changes the reference counter to 0 and detroys the A object. In main function after call to create, the reference counter should be of 1, and the object A should be available on the heap. But it has already been destroyed, leading to an incorrect result being displayed. In worst cases, a segmentation fault may happen. Remark : if changing the constructor of B, so that a const reference to the shared_ptr is passed, instead of the shared_ptr itself, it works : --- struct B { B (const shared_ptr <A> &val) : a (val) {} --- To further investigate the problem, it may be needed to analyse the assembly code generated by the gcc 4 compiler. The problem is clearly in the line : return D (C (B (a))); There may be a problem with management of temporary variables, combined with optimisations. I require the help of other developpers, because I think this problem could concern many of us, and not only boost users. * gcc 4.0.1 : Apple version coming with XCode 2.2.1 (build 5250) or gcc 4.0.2 delivered by darwinports. * boost library, tried with 1.33.1, 1.33.0 * using optimization level, either -O, -O1 or -O2 test_suite* init_unit_test_suite( int argc, char* argv[] ) { test_suite* test= BOOST_TEST_SUITE( "Sample test" ); boost::shared_ptr<SampleTest> instance (new SampleTest (12345)); test->add (BOOST_CLASS_TEST_CASE ( & SampleTest::sampleTest, instance)); return test; } ---- end-of-file boost_unit_test_gcc4_pb_sample.cpp --- And the command used for compiling : BOOST_PATH=<path-to-your-boost-install> /usr/bin/g++-4.0 -O2 -I${BOOST_PATH}/include -L${BOOST_PATH}/lib - lboost_unit_test_framework boost_unit_test_gcc4_pb_sample.cpp -o boost_unit_test_gcc4_pb_sample The result is the following : ---- Running 1 test case...a : 13691 boost_unit_test_gcc4_pb_sample.cpp(18): error in " SampleTest::sampleTest": check a == 12345 failed [13691 != 12345] It seems that when sampleTest () method is called by unit_test framework, adress of "a" variable is wrong, leading to a test failure. When, instead of using a class or struct attribute for variable "a", either a global, or class static variable is used, the problem disappears. Has anyone some experience using Boost unit_test framework on Mac OS X, using gcc 4 ? This email sent to site_archiver(a)lists.apple.com

1 0

Re: db bug with i386: sanity check
by Stephane Sudre 22 Mar '06

22 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 21 mars 06, at 19:17, Eric Albert wrote: On Mar 21, 2006, at 5:35 AM, Stephane Sudre wrote: Done. Actually I probably found where the problem exactly lies. It's not with the put API, it's with the sync API. If you do something like: open get put sync exit(0); It works fine on an any platform. If you do something like: open get put sync sync exit(0); _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… The more I look into this db thing, the more I believe there is a bug in the db code. So before filing a bug report, I may need a sanity check and hope there's a doctor in the room. It sounds like a bug to me. Go ahead and write it up. Thanks! It corrupts the data base on the Intel platform (PPC is fine). Doing 2 or more sync in the same execution is killing the data base. At least that's what I'm observing. This email sent to site_archiver(a)lists.apple.com

1 0

Re: db bug with i386: sanity check
by Eric Albert 21 Mar '06

21 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Mar 21, 2006, at 5:35 AM, Stephane Sudre wrote: It sounds like a bug to me. Go ahead and write it up. Thanks! Eric _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… The more I look into this db thing, the more I believe there is a bug in the db code. So before filing a bug report, I may need a sanity check and hope there's a doctor in the room. This email sent to site_archiver(a)lists.apple.com

1 0

db bug with i386: sanity check
by Stephane Sudre 21 Mar '06

21 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com --- Sanity check: ------------- [ ] Do you need to byte swap the key data on a put operation? a) Yes c) It depends on the lorder b) No d) If you are using a fixed-size ASCII string for the key, no. [ ] Do you need to byte swap the key length on a put operation? a) Yes c) Toss a coin. b) No d) It depends on the lorder. [ ] Do you need to byte swap the data length on a put operation? a) Yes c) It depends on the lorder b) No d) What is a swap? I tend to believe the answers are d / b / b. _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… The more I look into this db thing, the more I believe there is a bug in the db code. So before filing a bug report, I may need a sanity check and hope there's a doctor in the room. Let's say you have a HASH db base created with the lorder to be 0 on a PowerPC Mac OS X System. Inspection and update of the base works perfectly OK on a lovely PowerPC Mac. Let's say you inspect this base on an Intel Mac OS X System: it works perfectly. Let's say you replace an item in the base on an Intel Mac OS X System: it corrupts the base. This email sent to site_archiver(a)lists.apple.com

1 0

Re: S.M.A.R.T.-checking on Intel Macs impossible?
by Julian Mayer 21 Mar '06

21 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: S.M.A.R.T.-checking on Intel Macs impossible?
by Chris Sarcone 21 Mar '06

21 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: S.M.A.R.T.-checking on Intel Macs impossible?
by Julian Mayer 20 Mar '06

20 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0