1 ==================== Changes in man-pages-3.67 ====================
3 Released: ????-??-??, Munich
9 The following people contributed patches/fixes or (noted in brackets
10 in the changelog below) reports, notes, and ideas that have been
11 incorporated in changes in this release:
14 Apologies if I missed anyone!
17 New and rewritten pages
18 -----------------------
21 Michael Kerrisk, Peter Zijlstra [Juri Lelli]
22 New page describing sched_setattr(2) and sched_getattr(2)
26 Rewrote large parts of the page and added a number of details
29 Newly documented interfaces in existing pages
30 ---------------------------------------------
33 Peter Zijlstra, Michael Kerrisk [Juri Lelli]
34 Document SCHED_DEADLINE
35 Raw text from Peter Zijlstra
46 New link to new sched_setattr.
49 Changes to individual pages
50 ---------------------------
54 ERRORS: Add EADDRINUSE for ephemeral port range exhaustion
58 NOTES: Add some subheadings
61 Michael Kerrisk [William Morriss]
62 ERRORS: Add EADDRNOTAVAIL for ephemeral port range exhaustion
63 Verified from testing and the kernel source.
64 See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=745775
66 Remove mention of ip_local_port_range under EAGAIN error
67 Man-pages commit ac7477eb7a8 added EADDRNOTAVAIL as the correct
68 error for ephemeral port range exhaustion. However, that error
69 was also incorrectly described under the EAGAIN entry. So,
70 remove that text under EAGAIN.
76 Michael Kerrisk [Lucas De Marchi]
77 Clarify glibc header file declaration/ABI wrapper details
78 create_module(), delete_module(), init_module(), and
79 query_module() are not declared in header files, but
80 through an accident of history glibc provides an ABI
81 for them that it continues to maintain, for
82 compatibility reasons.
85 Michael Kerrisk [Steven Stewart-Gallus]
86 Note SIGKILL case when execve() fails beyond the point of no return
88 NOTES: Add a subheading and reorder paragraphs
91 Heinrich Schuchardt [Michael Kerrisk]
92 Document range of permitted flags for event_f_flags
93 With a new patch included in the mm tree, event_f_flags is
94 checked for allowable values.
98 Add "file locking" subheading under NOTES
102 ERRORS: SCHED_DEADLINE tasks can fail with EAGAIN
103 SCHED_DEADLINE tasks can fail with EAGAIN unless the
104 reset-on-fork flag is set.
108 Note that there is no glibc wrapper
112 Fix prototypes for getpriority() and setpriority()
113 The who argument has type id_t (which happens to be u32 on linux).
117 Add <sys/types.h> to synopsis of get_robust_list()
118 If one were to implement wrappers for [gs]et_robust_list() using the
119 given prototypes, one would also have to include sys/types.h to get
120 a definition of size_t.
123 Michael Kerrisk [Bill Allombert]
124 _GNU_SOURCE must be defined to obtain RUSAGE_THREAD definition
125 See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=746569
129 Andy Lutomirski [Michael Kerrisk]
130 Update AT_EMPTY_PATH and O_PATH documentation
134 ERRORS: Add EADDRINUSE for ephemeral port range exhaustion
138 Fix prototype for mbind(2)
139 The nmask argument is const. The return type in numaif.h is long.
140 (Well, at least <numaif.h> says nmask is const. The current kernel
141 does not define it as a const argument, but sys_mbind() only
142 passes it to get_nodes(), which does treat it as const.)
145 Davidlohr Bueso [Michael Kerrisk]
146 Enhance description of "full queue" criteria
150 Add <signal.h> to synopsis for ppoll()
151 One needs to #include <signal.h> to get the definition of the
152 type (sigset_t) of the mask argument to ppoll().
156 Fix return type of readlinkat()
160 Clarify details of msg_name and msg_namelen fields
165 See FIXME https://bugzilla.kernel.org/show_bug.cgi?id=75371
166 and http://thread.gmane.org/gmane.linux.man/5677
169 Andy Lutomirski [Christoph Hellwig, Andy Lutomirski]
170 remap_file_pages() has no benefit for real files
171 Linux commit 3ee6dafc677a68e461a7ddafc94a580ebab80735 caused
172 remap_file_pages to be emulated when used on real file.
174 sched_get_priority_max.2
176 'policy' can also be SCHED_DEADLINE
180 Fix prototype for sched_setaffinity()
181 The mask argument is const.
185 ERRORS: mark errors that apply just for sched_setparam()
187 ERRORS: Add EINVAL for invalid arguments
189 SEE ALSO: add sched_setattr(2)
193 ERRORS: mark errors that apply just to sched_setscheduler()
195 ERRORS: add EINVAL case for pid < 0
197 ERRORS: separate out EINVAL cases
201 NOTES: Add subheadings and reorder paragraphs
205 Fix prototypes for semop() and semtimedop()
206 The nsops arguments have type size_t, not unsigned, and the
207 timeout argument of semtimedop() is const.
209 NOTES: Add a subheading
213 Add details on various 'msghdr' fields
215 ERRORS: Add EAGAIN for ephemeral port range exhaustion
217 Add some subheadings under DESCRIPTION
221 NOTES: Add a subheading
224 Michael Kerrisk [Aurelien Jarno]
225 Describe feature test macro requirements for file type test macros
226 See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=728240
228 Update FTM requirements for lstat()
230 Split discussion of 'st_mode' fields into type and permissions
232 Move text on S_I*() macros to follow text on S_I* macros
233 That ordering is more logical
237 Fix prototype for stime()
238 The argument is const, both according to the actual header files and
239 according to <http://www.sco.com/developers/devspecs/vol1a.pdf>.
243 Fix prototype for syscall()
244 The first argument and the return value of syscall() has type long,
249 EXAMPLE: Add subheadings to distinguish the two example programs
253 Reword text referring to mallopt(3)
254 Linux libc is no longer "recent"; drop mention of it.
256 pthread_attr_setinheritsched.3
257 pthread_attr_setschedparam.3
258 pthread_attr_setschedpolicy.3
259 pthread_setaffinity_np.3
260 pthread_setschedparam.3
261 pthread_setschedprio.3
265 Change references to "sched_setscheduler(2)" to "sched(7)"
266 Change consistent with the fact that the scheduling overview
267 page is now sched(7) not sched_setscheduler(2).
269 pthread_attr_setstackaddr.3
271 ATTRIBUTES: Note functions that are thread-safe
272 The functions pthread_attr_setstackaddr() and
273 pthread_attr_getstackaddr() are thread safe.
275 pthread_attr_setstacksize.3
277 ATTRIBUTES: Note functions that are thread-safe
278 The functions pthread_attr_setstacksize() and
279 pthread_attr_getstacksize() are thread safe.
283 ATTRIBUTES: Note function that is thread-safe
284 The function pthread_kill() is thread safe.
286 pthread_kill_other_threads_np.3
288 ATTRIBUTES: Note function that is thread-safe
289 The function pthread_kill_other_threads_np() is thread safe.
293 ATTRIBUTES: Note function that is thread-safe
294 The function pthread_self() is thread safe.
296 pthread_setcancelstate.3
298 Add paragraph breaks to "Asynchronous cancelability" subsection
300 pthread_setcancelstate.3
302 ATTRIBUTES: Note functions that are thread-safe
303 The functions pthread_setcancelstate() and
304 pthread_setcanceltype() are thread safe.
306 NOTES: Add some subheadings
308 pthread_setschedparam.3
310 ATTRIBUTES: Note functions that are thread-safe
311 The functions pthread_setschedparam() and pthread_getschedparam()
314 pthread_setschedprio.3
316 ATTRIBUTES: Note function that is thread-safe
317 The function pthread_setschedprio() is thread safe.
321 ATTRIBUTES: Note function that is thread-safe
322 The function pthread_sigmask() is thread safe.
326 ATTRIBUTES: Note function that is thread-safe
327 The function pthread_sigqueue() is thread safe.
331 ATTRIBUTES: Note function that is thread-safe
332 The function pthread_testcancel() is thread safe.
336 ATTRIBUTES: Note function that is thread-safe
337 The function pthread_yield() is thread safe.
341 ATTRIBUTES: Note functions that are thread-safe
342 The functions remquo(), remquof() and remquol() are thread safe.
346 ATTRIBUTES: Note function that is thread-safe
347 The function rtime() is thread safe.
351 ATTRIBUTES: Note function that is thread-safe
352 The function sched_getcpu() is thread safe.
356 Note some history of stpcpy()
358 I just noticed that the stpcpy(3) manual contains a speculation
359 that appears to be untrue on closer investigation: That function
360 did not originate in MS DOS, but in Lattice C on AmigaDOS.
362 Here is a patch against the git master HEAD to fix that, and add
363 some more historical information. To provide some background and
364 allow you to more easily verify the correctness of the patch, i'm
365 appending my mail to <misc@openbsd.org>, where i'm giving some
366 more details about the history and pointing to some primary
367 sources. That mail also contains the (similar, but shorter)
368 patch i just committed to the OpenBSD manual page.
371 Michael Kerrisk [Aurelien Jarno, Török Edwin]
372 Explain why strcasecmp()+strncasecmp() are also declared in <string.h>
373 See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=729436
377 NOTES: Add a subheading
380 Michael Kerrisk [Sam Varshavchik]
381 Fix floppy disk device names
382 The naming convention shown in the page was ancient.
383 Now, the page is consistent with Documentation/devices.txt
384 (where it is noted that "The use of the capital letters
385 D, H and E for the 3.5" models have been deprecated, since
386 the drive type is insignificant for these devices"
390 Document /proc/timer_stats
392 (Briefly) document /proc/timer_list
393 Text taken from kernel commit
394 289f480af87e45f7a6de6ba9b4c061c2e259fe98.
396 Add /proc/sys/kernel/{sched_rt_period_us,sched_rt_runtime_us}
397 Reference sched(7) for an explanation of these two files
401 Mention sched_setattr(2) under CAP_SYS_NICE
405 SEE ALSO: add sched(7)
409 Mention sched_getattr() as a place where credentials are checked
412 Heinrich Schuchardt [Jan Kara]
413 BUGS: error events can be lost when reading from fanotify FD
414 Michael Kerrisk [Heinrich Schuchardt]
415 Fix description of FAN_EVENT_NEXT()
416 FAN_EVENT_NEXT() does not update 'meta'; rather, it returns a
417 pointer to the next metadata structure. In addition, generally
418 rework the description to be a bit clearer and more detailed.
420 Document FAN_EVENT_METADATA_LEN
424 Note cases where an ephemeral port is used
426 Remove BUGS text on glibc failing to declare in_pktinfo
428 Clarify 'ip_local_port_range' and mention the term "ephemeral ports"
430 Note some more details about assignment of ephemeral ports
432 BUGS: ephemeral port range exhaustion is diagnosed inconsistently
433 Different system calls use different 'errno' values to diagnose
434 exhaustion of the ephemeral port range.
438 Document sched_rt_period_us and sched_rt_runtime_us /proc files
439 And rework and relocate the text on dealing with runaway
442 Mention sched_setattr(2) in list of APIs that can change policies
444 sched_setattr(2) can also be used to set 'nice' value
446 Remove mention of sched_setscheduler() when talking about sched_priority
447 There are nowadays multiple ways to set sched_priority (and
448 in fact there always were, since we also had sched_setparam(2)).
450 SEE ALSO: Add Documentation/scheduler/sched-design-CFS.txt
452 Don't mention sched_setscheduler(2) in discussions of setting policies
453 In a couple of places, sched_setscheduler(2) is mentioned as the
454 way of setting policies. But now there is sched_setattr(2) as
455 well, rewrite the text in a more generic way.
457 Rework summary text describing sched_setattr(2) and sched_getattr(2)
458 Note that these APIs are a superset of sched_setscheduler(2)
459 and sched_getscheduler(2).
461 Remove crufty text relating to sched_setscheduler()
462 All of the removed text is in sched_setscheduler(2) and
463 should have been trimmed from this page.
465 SEE ALSO: Mention more files in Documentation/scheduler/ directory