[Thread] TSS + static linking status
What's the status of TSS cleanup issue in static library configuration for the following compilers: CodeWarrior 8.3 MinGW GCC 3.x Borland 5.6.4 Intel C++ 7.1 (for Windows) ? Should the corresponding test's failures (e.g. http://tinyurl.com/6v6sd) be marked as expected? -- Aleksey Gurtovoy MetaCommunications Engineering
Aleksey Gurtovoy wrote:
What's the status of TSS cleanup issue in static library configuration for the following compilers:
MinGW GCC 3.x
I'm the only person who has volunteered to get it working on this platform, and unfortunately, there is no quick fix possible here. I have separately been working on some additions to binutils and mingwrt that would make static TSS possible; but this will not be integrated into MinGW in time for the Boost release. Aaron W. LaFramboise
Aaron W. LaFramboise writes:
Aleksey Gurtovoy wrote:
What's the status of TSS cleanup issue in static library configuration for the following compilers:
MinGW GCC 3.x
I'm the only person who has volunteered to get it working on this platform, and unfortunately, there is no quick fix possible here. I have separately been working on some additions to binutils and mingwrt that would make static TSS possible; but this will not be integrated into MinGW in time for the Boost release.
Aaron, Thanks for the information. I'm going to mark the corresponding failure as expected, then, with the following note: This functionality has not been implemented yet. The library developers plan to implement it for the next release. -- Aleksey Gurtovoy MetaCommunications Engineering
"Aleksey Gurtovoy" <agurtovoy@meta-comm.com> wrote in message news:uy8j5siui.fsf@meta-comm.com...
Aaron W. LaFramboise writes:
Aleksey Gurtovoy wrote:
What's the status of TSS cleanup issue in static library configuration for the following compilers:
MinGW GCC 3.x
I'm the only person who has volunteered to get it working on this platform, and unfortunately, there is no quick fix possible here. I have separately been working on some additions to binutils and mingwrt that would make static TSS possible; but this will not be integrated into MinGW in time for the Boost release.
Aaron,
Thanks for the information. I'm going to mark the corresponding failure as expected, then, with the following note:
This functionality has not been implemented yet. The library developers plan to implement it for the next release.
Yes, this is what should be done. Thanks. Mike
Michael Glassford writes:
"Aleksey Gurtovoy" <agurtovoy@meta-comm.com> wrote in message news:uy8j5siui.fsf@meta-comm.com...
Aaron W. LaFramboise writes:
Aleksey Gurtovoy wrote:
What's the status of TSS cleanup issue in static library configuration for the following compilers:
MinGW GCC 3.x
I'm the only person who has volunteered to get it working on this platform, and unfortunately, there is no quick fix possible here. I have separately been working on some additions to binutils and mingwrt that would make static TSS possible; but this will not be integrated into MinGW in time for the Boost release.
Aaron,
Thanks for the information. I'm going to mark the corresponding failure as expected, then, with the following note:
This functionality has not been implemented yet. The library developers plan to implement it for the next release.
Yes, this is what should be done. Thanks.
For the rest of the list: CodeWarrior 8.3 Borland 5.6.4 Intel C++ 7.1 (for Windows) as well, right? -- Aleksey Gurtovoy MetaCommunications Engineering
Michael Glassford writes:
"Aleksey Gurtovoy" <agurtovoy@meta-comm.com> wrote in message news:uy8j5siui.fsf@meta-comm.com...
Aaron W. LaFramboise writes:
Aleksey Gurtovoy wrote:
What's the status of TSS cleanup issue in static library configuration for the following compilers:
MinGW GCC 3.x
I'm the only person who has volunteered to get it working on
"Aleksey Gurtovoy" <agurtovoy@meta-comm.com> wrote in message news:cimkfi$pr4$1@sea.gmane.org... this
platform, and unfortunately, there is no quick fix possible here. I have separately been working on some additions to binutils and mingwrt that would make static TSS possible; but this will not be integrated into MinGW in time for the Boost release.
Aaron,
Thanks for the information. I'm going to mark the corresponding failure as expected, then, with the following note:
This functionality has not been implemented yet. The library developers plan to implement it for the next release.
Yes, this is what should be done. Thanks.
For the rest of the list:
CodeWarrior 8.3
Yes.
Borland 5.6.4
Yes.
Intel C++ 7.1 (for Windows)
I'm not sure. Since Intel 8 passes, it would be good to see if Intel 7.1 could pass too.
as well, right?
I believe CW9.3 should be marked as well. It's actually passing for some reason, but I don't understand how it could; Stefan Slapeta is looking into it. I think it would be best to mark it as an expected failure so that the results show that its passing the tests is unexpected, at least until we figure out that why. Possibly also VC7. VC6 and VC7.1 are supported, but I don't believe anyone ever looked into whether VC7 worked. Since the regression test is failing, it looks like either additional work is needed to support it or it should be marked as an expected failure for now. Thanks, Mike
Michael Glassford writes:
"Aleksey Gurtovoy" <agurtovoy@meta-comm.com> wrote in message
Aaron,
Thanks for the information. I'm going to mark the corresponding failure as expected, then, with the following note:
This functionality has not been implemented yet. The library developers plan to implement it for the next release.
Yes, this is what should be done. Thanks.
For the rest of the list:
CodeWarrior 8.3
Yes.
Borland 5.6.4
Yes.
Intel C++ 7.1 (for Windows)
I'm not sure. Since Intel 8 passes, it would be good to see if Intel 7.1 could pass too.
I've marked it as expected for the time being. If somebody fixes it before the release, I'll remove the markup.
as well, right?
I believe CW9.3 should be marked as well. It's actually passing for some reason, but I don't understand how it could; Stefan Slapeta is looking into it. I think it would be best to mark it as an expected failure so that the results show that its passing the tests is unexpected, at least until we figure out that why.
OK.
Possibly also VC7. VC6 and VC7.1 are supported, but I don't believe anyone ever looked into whether VC7 worked. Since the regression test is failing, it looks like either additional work is needed to support it or it should be marked as an expected failure for now.
OK, everything is marked up now according to your comments. The reports should reflect it within two hours. Thanks for the assistance, -- Aleksey Gurtovoy MetaCommunications Engineering
"Aleksey Gurtovoy" <agurtovoy@meta-comm.com> wrote in message news:m2d60gv0xr.fsf@meta-comm.com... [snip]
OK, everything is marked up now according to your comments. The reports should reflect it within two hours.
Thanks for the assistance,
Thanks for doing it. I've been meaning to figure out how to do it but hadn't found the time. Mike
Michael Glassford wrote:
I believe CW9.3 should be marked as well. It's actually passing for some reason, but I don't understand how it could; Stefan Slapeta is looking into it. I think it would be best to mark it as an expected failure so that the results show that its passing the tests is unexpected, at least until we figure out that why.
Bad news: also the results for Intel are incorrect, the static cleanup doesn't work for Intel Compiler, either: Running 1 test case... tss_instances = 0; tss_total = 5 tss_instances = 5; tss_total = 5 ../libs/thread/test/test_tss.cpp(152): error in "test_tss": test tss_instances == 0 failed [5 != 0] *** 1 failure detected in test suite "Boost.Threads: tss test suite" EXIT STATUS: 0 I'll markup this as expected failure and submit a feature request for the Intel Compiler. Stefan
Stefan Slapeta wrote:
I'll markup this as expected failure and submit a feature request for the Intel Compiler.
I've commited an update for the thread tss cleanup which makes it working for Intel, too. The feature was just not activated for this compiler :( [Once again is was the BOOST_MSVC macro, which is set _only_ for VC and _not_ for Intel for Windows.] I've just tested Intel 8, don't know how version 7 will behave. We'll see in the regression results. Stefan
"Stefan Slapeta" <stefan@slapeta.com> wrote in message news:ciud1n$tla$1@sea.gmane.org...
Stefan Slapeta wrote:
I'll markup this as expected failure and submit a feature request
for
the Intel Compiler.
I've commited an update for the thread tss cleanup which makes it working for Intel, too. The feature was just not activated for this compiler :(
Good news! [snip] Mike
Michael Glassford wrote:
"Stefan Slapeta" wrote in message
I've commited an update for the thread tss cleanup which makes it working for Intel, too. The feature was just not activated for this compiler :(
Good news!
Seems to work for Intel 7, too! Stefan
Michael Glassford wrote:
I believe CW9.3 should be marked as well. It's actually passing for some reason, but I don't understand how it could; Stefan Slapeta is looking into it. I think it would be best to mark it as an expected failure so that the results show that its passing the tests is unexpected, at least until we figure out that why.
When I execute the test manually, it fails! Output (as expected): test_tss.cpp(152): error in "test_tss": test tss_instances == 0 failed [5 != 0] However, below is what I can find in bjam.log for test_tss_lib. It seems that the test never gets executed. Any explanation? Stefan execute-test C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.run set PATH=c:\Programme\Metrowerks\Codewarrior\Win32-x86 Support\Libraries\Runtime;c:\Programme\Metrowerks\Codewarrior\Win32-x86 Support\Libraries\Runtime\Libs\MSL_All-DLLs;%PATH% C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.exe > C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.output 2>&1 set status=%ERRORLEVEL% echo. >> C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.output echo EXIT STATUS: %status% >> C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.output if %status% EQU 0 ( copy C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.output C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.run ) set verbose=1 if %status% NEQ 0 ( set verbose=0 ) if %verbose% EQU 0 ( echo ====== BEGIN OUTPUT ====== type C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.output echo ====== END OUTPUT ====== ) exit %status% 1 Datei(en) kopiert. RmTemps C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.run del /f/q "C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.exe" **passed** C:\boost_regression\results\bin\boost\libs\thread\test\test_tss_lib.test\cw-9.3\debug\runtime-link-static\threading-multi\test_tss_lib.test
participants (4)
-
Aaron W. LaFramboise -
Aleksey Gurtovoy -
Michael Glassford -
Stefan Slapeta