Discussions of the Parallel Programming book
 help / color / mirror / Atom feed
* POSIX Proess Creation and Destruction nit-picking
@ 2019-04-26 11:23 Elad Lahav
  2019-04-27 16:19 ` Akira Yokosawa
  0 siblings, 1 reply; 6+ messages in thread
From: Elad Lahav @ 2019-04-26 11:23 UTC (permalink / raw)
  To: perfbook

The following issues may not be deemed important enough to fix, but
I'll point them out nevertheless:

- kill() does not destroy a process, but merely sends a signal. The
action taken by the process upon the delivery of the signal depends on
the signal itself, whether it is currently masked, and whether the
process has installed a signal handler. The only signal guaranteed to
destroy a process is SIGKILL, which may not be masked and cannot be
associated with a handler.
- "It is critically important to note that the parent and child do not
share memory." - That is not true. Parent and child share all memory
marked as shared (i.e., mappings created with mmap(MAP_SHARED)). The
example works because by default the BSS is mapped as private.
- Listing 4.2 could have just used waitpid() instead of the loop.

If I were really a nit-picking person I would have pointed out the
various places in chapters 2 and 3 where infinitives are being split,
but I'm not so I won't ;-).

--Elad


^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2019-05-01 17:15 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-04-26 11:23 POSIX Proess Creation and Destruction nit-picking Elad Lahav
2019-04-27 16:19 ` Akira Yokosawa
2019-04-27 23:00   ` Akira Yokosawa
2019-04-28  1:54     ` Elad Lahav
2019-04-30  0:12   ` [PATCH v2] toolsoftrade: Fix extraction range of waitall() Akira Yokosawa
2019-05-01 17:15     ` Paul E. McKenney

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox