Bruce Korb wrote: > "implementations" are now allowed to reset the handler to > SIG_DFL on exec(2)
Ah, an improvement indeed. > so it is "may not work" not "will not work". No, it's still "will not work" because the text is talking about the setting of SIGCHLD in the process that it calling wait_subprocess. The setting in the child process has no effect here, and the setting in the parent process is not directly relevant since it may have been corrected in the current process. Bruno
