Thread/intel-win32-8.1 failures
As shown in http://tinyurl.com/6ukqm, _all_ thread library tests fail on Intel 8.1 with memory access violation (!). That's especially weird because Intel 8.0 shows quite the opposite results -- it passes all the tests. Could somebody with local access to this compiler please take a look at this? -- Aleksey Gurtovoy MetaCommunications Engineering
Hi Aleksey,
As shown in http://tinyurl.com/6ukqm, _all_ thread library tests fail on Intel 8.1 with memory access violation (!). That's especially weird because Intel 8.0 shows quite the opposite results -- it passes all the tests.
see http://tinyurl.com/62v2b. Nothing we can do without a new compiler version... Stefan
Stefan Slapeta writes:
As shown in http://tinyurl.com/6ukqm, _all_ thread library tests fail on Intel 8.1 with memory access violation (!). That's especially weird because Intel 8.0 shows quite the opposite results -- it passes all the tests.
see http://tinyurl.com/62v2b. Nothing we can do without a new compiler version...
Should we mark the toolset as unusable with a corresponding comment, then? We can always remove the markup when/if it gets fixed. -- Aleksey Gurtovoy MetaCommunications Engineering
Aleksey Gurtovoy wrote:
Should we mark the toolset as unusable with a corresponding comment, then? We can always remove the markup when/if it gets fixed.
The problem is not in the thread library but in boost.test, where the address of a virtual function is not taken properly. The error occurs whenever the program execution monitor is used. What makes sense IMO is that I should disable the Intel 8.1 tests until a compiler update is released because they have no value for us at the current state. Stefan
Stefan Slapeta writes:
Aleksey Gurtovoy wrote:
Should we mark the toolset as unusable with a corresponding comment, then? We can always remove the markup when/if it gets fixed.
The problem is not in the thread library but in boost.test, where the address of a virtual function is not taken properly. The error occurs whenever the program execution monitor is used.
Oh, OK.
What makes sense IMO is that I should disable the Intel 8.1 tests until a compiler update is released because they have no value for us at the current state.
Let's do that. -- Aleksey Gurtovoy MetaCommunications Engineering
Aleksey Gurtovoy wrote:
What makes sense IMO is that I should disable the Intel 8.1 tests until a compiler update is released because they have no value for us at the current state.
Let's do that.
Tests for Intel 8.1 are stopped. John, do you know anything more on when the update for the 8.1 compiler is due to be released? Do you think you can get earlier updates? At the moment I don't know what they are planning to do because I don't get much response to my latest issues... Stefan
Tests for Intel 8.1 are stopped.
John, do you know anything more on when the update for the 8.1 compiler is due to be released? Do you think you can get earlier updates?
I don't think so, all I know is that when I raised this with developer relations, they said that the issue was fixed in their current compiler build, but that this wouldn't make the initial 8.1 release. So I'm assuming it'll make the next release... they usually seem to come out quite often (monthly?). John.
participants (3)
-
Aleksey Gurtovoy -
John Maddock -
Stefan Slapeta