[test] SIGPOLL not defined on FreeBSD?
If possible I'd like to see a fix for: http://tinyurl.com/3uz5xg Looks like we just need to add FreeBSD to the list of OS's that don't define SIGPOLL at the start of execution_monitor.ipp. The strange thing is, the other FreeBSD tester doesn't fail here... anyone know what the difference is between them? Both define XOPEN_UNIX to be -1 so they're both advertising that they *don't* support this XSI-specific feature. Come to that are these symbols all macros? Could we just check for defined(SIGPOLL) ? Cheers, John.
John Maddock wrote:
If possible I'd like to see a fix for: http://tinyurl.com/3uz5xg
Looks like we just need to add FreeBSD to the list of OS's that don't define SIGPOLL at the start of execution_monitor.ipp. The strange thing is, the other FreeBSD tester doesn't fail here... anyone know what the difference is between them? Both define XOPEN_UNIX to be -1 so they're both advertising that they *don't* support this XSI-specific feature. Come to that are these symbols all macros? Could we just check for defined(SIGPOLL) ?
Cheers, John.
I'll look into this once I install and configure vmware image. Gennadiy
Hello I can confirm this. I had tried to build boost 1.35 on freebsd 6.2 and found this bug. There is also some problems with suboptions to SIGCHLD and other signals - i can look for log at my vmware.
"JM" == John Maddock writes: JM> If possible I'd like to see a fix for: http://tinyurl.com/3uz5xg Looks JM> like we just need to add FreeBSD to the list of OS's that don't define JM> SIGPOLL at the start of execution_monitor.ipp. The strange thing is, JM> the other FreeBSD tester doesn't fail here... anyone know what the JM> difference is between them? Both define XOPEN_UNIX to be -1 so JM> they're both advertising that they *don't* support this XSI-specific JM> feature. Come to that are these symbols all macros? Could we just JM> check for defined(SIGPOLL) ?
-- With best wishes, Alex Ott, MBA http://alexott.blogspot.com/ http://xtalk.msk.su/~ott/ http://alexott-ru.blogspot.com/
Alex Ott <alexott <at> gmail.com> writes:
Hello
I can confirm this. I had tried to build boost 1.35 on freebsd 6.2 and found this bug. There is also some problems with suboptions to SIGCHLD and other signals - i can look for log at my vmware.
SIGPOLL issue fix is checked in yesterday. The rest is very specific to FreeBSD version before 7 and I am a bit reluctant to invest time in unless it really critical. If you are interested please submit the patch. Gennadiy
participants (3)
-
Alex Ott -
Gennadiy Rozental -
John Maddock