[Bug 191744] lang/python27: With THREADS option: devel/pth: pthread.h conflicts with system pthread.

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191744

--- Comment #28 from mikhail.rokhin at gmail.com ---

 ┌──────────────────────────── gnupg-2.0.25 ──────────────────────────────┐
  │ ┌────────────────────────────────────────────────────────────────────┐ │  
  │ │ [ ] PINENTRY    Use pinentry                                       │ │  
  │ │ [ ] LDAP        LDAP keyserver interface                           │ │  
  │ │ [ ] SCDAEMON    Enable Smartcard daemon (with libusb)              │ │  
  │ │ [x] CURL        Use the real curl library (worked around if no)    │ │  
  │ │ [ ] GPGSM       Enable GPGSM (requires LDAP)                       │ │  
  │ │ [ ] KDNS        Use DNS CERT helper                                │ │  
  │ │ [ ] STD_SOCKET  Use standard socket for agent                      │ │  
  │ │ [x] NLS         Native Language Support                            │ │  
  │ │ [x] DOCS        Build and/or install documentation                 │ │  
  │ │ [ ] SUID_GPG    Install GPG with suid                              │ │  
  │ └────────────────────────────────────────────────────────────────────┘ │  
  ├────────────────────────────────────────────────────────────────────────┤  
  │                     <  OK  >           <Cancel>                        │  
  └────────────────────────────────────────────────────────────────────────┘  


(In reply to mikhail.rokhin from comment #27)
> Finally found what installed it
> 
> devel/pth: security/gnupg
> 
> 
> (In reply to Jung-uk Kim from comment #14)
> > It looked at the problem more carefully.  This issue is little complicated. 
> > Basically, sys/types.h includes sys/_pthreadtypes.h per POSIX since r146824.
> > 
> > http://svnweb.freebsd.org/changeset/base/146824
> > 
> > Unfortunately, devel/pth tries to override the pthread types defined in
> > sys/_pthreadtypes.h via its own pthread.h.  If sys/types.h is also included,
> > these types get redefined.  Unfortunately, this problem is really hard to
> > avoid.

-- 
You are receiving this mail because:
You are the assignee for the bug.