API tools faq
paste
Advertisement
Guest User

Untitled

a guest
Sep 24th, 2015
134
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 34.77 KB | None | 0 0
raw download clone embed print report
  1. # git clone https://gitlab.com/procps-ng/procps.git
  2. # cd procps
  3. # git log --author=warner
  4. commit b4923fa7450824f9ed85fe5ce0d7a0d721329304
  5. Author: Jim Warner <james.warner@comcast.net>
  6. Date: Thu Aug 6 08:43:48 2015 -0500
  7.  
  8. miscellaneous: a final cleanup prior to release 3.3.11
  9.  
  10. This patch just eliminates some eol whitespace, adds a
  11. missing eof newline and contributes yet one additional
  12. entry to the NEWS summary regarding saved top rcfiles.
  13.  
  14. Signed-off-by: Jim Warner <james.warner@comcast.net>
  15.  
  16. commit 3da298c1d6b9c82e99e7025162657a8b5e2ce551
  17. Author: Jim Warner <james.warner@comcast.net>
  18. Date: Mon Aug 3 00:00:00 2015 -0500
  19.  
  20. top: fix unlikely edge case wherein all fields are off
  21.  
  22. While testing a newlib interface for pids acquisitions
  23. I encountered some unexpected results if an idiot user
  24. (me) turns off all displayable fields. So, this commit
  25. ensures that the PID field will be shown as a minimum.
  26.  
  27. Signed-off-by: Jim Warner <james.warner@comcast.net>
  28.  
  29. commit 8226ca983856d2d023dc52e5d56ec92552b5bf18
  30. Author: Jim Warner <james.warner@comcast.net>
  31. Date: Fri Jul 31 00:00:00 2015 -0500
  32.  
  33. top: avoid an unnecessary conversion for 'USED' column
  34.  
  35. When the USED column was introduced the proc_t.vm_swap
  36. & proc_t.resident values were added together. However,
  37. using 'resident' required an additional PROC_FILL flag
  38. not to mention extra conversion of pages to kibibytes.
  39.  
  40. So now we'll use an already present vm_rss value which
  41. removes any special handling for top's derived column.
  42.  
  43. And while we're at it we'll trade some more 'resident'
  44. field uses with that more immediately usable 'vm_rss'.
  45.  
  46. Reference(s):
  47. commit 709785e20bd19dc28546d19c45bb7444a56f88b9
  48.  
  49. Signed-off-by: Jim Warner <james.warner@comcast.net>
  50.  
  51. commit 9ba65bad8ef1bb994f3edd45063878daa1beb78c
  52. Author: Jim Warner <james.warner@comcast.net>
  53. Date: Mon Jul 27 00:00:00 2015 -0500
  54.  
  55. top: miscellaneous accumulated tweaks to code/comments
  56.  
  57. Jeeze, to correct spelling on one single word (incure)
  58. you had to go and align the entire comments paragraph?
  59.  
  60. [ well, at least there's one other minor code change ]
  61.  
  62. Signed-off-by: Jim Warner <james.warner@comcast.net>
  63.  
  64. commit 9cc6ed10c87ac25ded3bcc1ad5c54d05d656d99e
  65. Author: Jim Warner <james.warner@comcast.net>
  66. Date: Mon Jul 27 00:00:00 2015 -0500
  67.  
  68. top: eliminate 'user' from the inspection view headers
  69.  
  70. Since it's possible that euser name is not being shown
  71. or the horizontal position had been scrolled past that
  72. USER column, then part of those headers will be blank.
  73.  
  74. So it doesn't make sense to try and show the USER that
  75. is associated with a process at all. Thus, this commit
  76. simply removes the 'user' provision from both headers.
  77.  
  78. Signed-off-by: Jim Warner <james.warner@comcast.net>
  79.  
  80. commit c07f6c5e6d6170ef3f2373b4d86988184c3235bf
  81. Author: Jim Warner <james.warner@comcast.net>
  82. Date: Sun Jul 12 04:44:44 2015 -0500
  83.  
  84. top: improve vertical scroll management for 'i' toggle
  85.  
  86. When a user is taking advantage of the scroll features
  87. it is likely a scrolled vertical position is well past
  88. the first displayable task. That is especially true of
  89. top's forest view ('V') mode where those early systemd
  90. attached processes are generally not very interesting.
  91.  
  92. As such, should the idle mode toggle ('i') be employed
  93. a distorted display is almost guaranteed because tasks
  94. that have used some cpu, and thus should be displayed,
  95. have already been skipped by virtue of their position.
  96.  
  97. So this patch temporarily nullifies vertical scrolling
  98. during the period when idle tasks are not being shown.
  99.  
  100. Signed-off-by: Jim Warner <james.warner@comcast.net>
  101.  
  102. commit 9e85c147b2ac4eb4177d87bab541f00f958e5a8d
  103. Author: Jim Warner <james.warner@comcast.net>
  104. Date: Sat Jun 20 00:00:00 2015 -0500
  105.  
  106. build-sys: tweak to prevent dirty tree with every make
  107.  
  108. Please let's stop the nls translation insanity. With a
  109. one time push we can eliminate the dirty tree syndrome
  110. which surfaces with every local build. Later, before a
  111. release, the translations can be updated in final form
  112. then pushed just 1 more time to the gitlab repository.
  113.  
  114. I'm tired of having to always re-issue this request in
  115. order to circumvent the problem and thus prevent a too
  116. broad commit (not to mention some nasty side effects).
  117.  
  118. [ bash$ git update-index --assume-unchanged po/??.po ]
  119.  
  120. Reference(s):
  121. http://www.freelists.org/post/procps/procpsng-translations
  122.  
  123. Signed-off-by: Jim Warner <james.warner@comcast.net>
  124.  
  125. commit 88fe45ef6815d7d06cc7efba70351fdc80277996
  126. Author: Jim Warner <james.warner@comcast.net>
  127. Date: Sat Jun 20 00:00:00 2015 -0500
  128.  
  129. top: sacrifice a redundant procps_linux_version() call
  130.  
  131. With the commit referenced below, the linux version is
  132. no longer available via an external variable. So we'll
  133. eliminate the extra superficial function call employed
  134. at program end as part of a debugging (only) o/p spew.
  135.  
  136. [ the user will soon be returned to the command line ]
  137. [ & he/she can run their own 'uname -r' if in doubt! ]
  138.  
  139. Reference(s):
  140. commit 56d9d5e7e701e71646679b2b8cefa9239ee59592
  141.  
  142. Signed-off-by: Jim Warner <james.warner@comcast.net>
  143.  
  144. commit 4129c758ac2eae553450bd1875942677618ff2fa
  145. Author: Jim Warner <james.warner@comcast.net>
  146. Date: Fri Jun 19 08:41:51 2015 -0500
  147.  
  148. miscellaneous: remove some trailing whitespace buildup
  149.  
  150. wish folks (craig) would use these in their .gitconfig
  151.  
  152. [core]
  153. whitespace = trailing-space, space-before-tab, blank-at-eof
  154. [apply]
  155. whitespace = warn
  156.  
  157. Signed-off-by: Jim Warner <james.warner@comcast.net>
  158.  
  159. commit ad0a0fc138e7f9520aac07966f2dd7830813497d
  160. Author: Jim Warner <james.warner@comcast.net>
  161. Date: Fri Jun 19 08:34:12 2015 -0500
  162.  
  163. build-sys: fix oversight for a newly added header file
  164.  
  165. Signed-off-by: Jim Warner <james.warner@comcast.net>
  166.  
  167. commit 3600f652e83b5b9e084e0a9ffecaeae4115d4fab
  168. Author: Jim Warner <james.warner@comcast.net>
  169. Date: Thu Jun 18 00:00:00 2015 -0500
  170.  
  171. documentation: fix man pages due to refactor for wchan
  172.  
  173. This patch will bring three of our man pages into line
  174. with the recent refactor of the libprocps wchan logic.
  175.  
  176. [ and also eliminates more damn eol whitespace which ]
  177. [ snuck in our repo with the commit referenced below ]
  178.  
  179. Reference(s):
  180. http://www.freelists.org/post/procps/WCHAN,11
  181. commit cf4788c28dba784a09c8b036a56fffa49da584e7
  182.  
  183. Signed-off-by: Jim Warner <james.warner@comcast.net>
  184.  
  185. commit 93666da62c3ef333a546f8a208c50e21d01107cd
  186. Author: Jim Warner <james.warner@comcast.net>
  187. Date: Thu Jun 18 00:00:00 2015 -0500
  188.  
  189. top: adapt to a simplified library interface for wchan
  190.  
  191. This patch was made necessary by those library changes
  192. in support of recently revised/simplified wchan logic.
  193.  
  194. Reference(s):
  195. http://www.freelists.org/post/procps/WCHAN,11
  196.  
  197. Signed-off-by: Jim Warner <james.warner@comcast.net>
  198.  
  199. commit caeebdc71f4def59af908e4afcf9b63e781fe4c6
  200. Author: Jim Warner <james.warner@comcast.net>
  201. Date: Thu Jun 18 00:00:00 2015 -0500
  202.  
  203. ps: exploit the simplified library interface for wchan
  204.  
  205. This patch was made necessary by those library changes
  206. in support of recently revised/simplified wchan logic.
  207.  
  208. In addition, this commit eliminates a broken alternate
  209. 'namelist' provision which was intended to allow users
  210. to specify a System.map file to be used in translating
  211. addresses into function names. But, the real effect of
  212. the now defunct 'N' and '-n' options was to indirectly
  213. force addresses (not names) to be displayed since such
  214. user named map files could not be successfully parsed.
  215.  
  216. Besides when the required FRAME_POINTER kconfig option
  217. is absent there is no address to translate and when it
  218. is present /proc/PID/wchan is already translated. Thus
  219. an alternate mapping is unnecessary and inappropriate.
  220.  
  221. [ we'll forgive POSIX for documenting '-n namelist' ]
  222.  
  223. Reference(s):
  224. http://www.freelists.org/post/procps/WCHAN,11
  225.  
  226. Signed-off-by: Jim Warner <james.warner@comcast.net>
  227.  
  228. commit 6b8dc5511fb9780806c3e8b97e5514ec30c2f514
  229. Author: Jim Warner <james.warner@comcast.net>
  230. Date: Thu Jun 18 00:00:00 2015 -0500
  231.  
  232. library: refactor and rely on modern kernels for wchan
  233.  
  234. Several Debian based distributions were recently found
  235. to have omitted a kernel configuration option that had
  236. the effect of rendering /proc/#/stat and /proc/#/wchan
  237. useless for providing any 'sleeping in function' info.
  238.  
  239. That problem also prompted a reevaluation of the whole
  240. approach to wchan matters which had grown increasingly
  241. complex as our library evolved over the last 13 years.
  242.  
  243. The net result was a decision to rely on /proc/#/wchan
  244. which arrived along with the 2.5 kernel. This then let
  245. us vastly simplify the internal code plus the external
  246. interface which will benefit both the top and ps pgms.
  247.  
  248. Reference(s):
  249. http://www.freelists.org/post/procps/WCHAN,11
  250. https://lkml.org/lkml/2008/11/6/12
  251. https://bugs.debian.org/711592
  252.  
  253. Signed-off-by: Jim Warner <james.warner@comcast.net>
  254.  
  255. commit 932f54b19d6b53d2dc804b291293052e7876fbda
  256. Author: Jim Warner <james.warner@comcast.net>
  257. Date: Mon Jun 15 00:00:00 2015 -0500
  258.  
  259. top: correct an alphabetic field error in man document
  260.  
  261. Signed-off-by: Jim Warner <james.warner@comcast.net>
  262.  
  263. commit 69580f7d98debb5df4b2050d30a1b027f6836838
  264. Author: Jim Warner <james.warner@comcast.net>
  265. Date: Tue May 12 03:00:00 2015 -0500
  266.  
  267. build-sys: do not distribute the aging README.top file
  268.  
  269. Signed-off-by: Jim Warner <james.warner@comcast.net>
  270.  
  271. commit 65773719196a5fa0d2aadf984821f256cc0ea275
  272. Author: Jim Warner <james.warner@comcast.net>
  273. Date: Sat Jun 13 00:00:00 2015 -0500
  274.  
  275. ps: update man document to support LXC container names
  276.  
  277. Signed-off-by: Jim Warner <james.warner@comcast.net>
  278.  
  279. commit f9e27afeb2ca28292223ab17395a546529d9704f
  280. Author: Jim Warner <james.warner@comcast.net>
  281. Date: Sat Jun 13 00:00:00 2015 -0500
  282.  
  283. ps: add code to exploit the new library LXC containers
  284.  
  285. Reference(s):
  286. https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/1424253
  287. https://bugs.launchpad.net/ubuntu/+source/procps/+bug/1424253
  288.  
  289. Signed-off-by: Jim Warner <james.warner@comcast.net>
  290.  
  291. commit d8aee1a809fdb4764432ed47f69e9f4bf4ce6778
  292. Author: Jim Warner <james.warner@comcast.net>
  293. Date: Sat Jun 13 00:00:00 2015 -0500
  294.  
  295. top: update man document to support the LXC containers
  296.  
  297. [ in addition to the primary 'lxc' business, i found ]
  298. [ numerous apostrophes used instead of that back-tic ]
  299.  
  300. Signed-off-by: Jim Warner <james.warner@comcast.net>
  301.  
  302. commit 2ba7aa8b7d4e1731044844589f736c5625a0b15e
  303. Author: Jim Warner <james.warner@comcast.net>
  304. Date: Sat Jun 13 00:00:00 2015 -0500
  305.  
  306. top: add support exploiting new library LXC containers
  307.  
  308. Reference(s):
  309. https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/1424253
  310. https://bugs.launchpad.net/ubuntu/+source/procps/+bug/1424253
  311.  
  312. Signed-off-by: Jim Warner <james.warner@comcast.net>
  313.  
  314. commit 0557504f9cb84987f9d9038755404be017bdb7d1
  315. Author: Jim Warner <james.warner@comcast.net>
  316. Date: Sat Jun 13 00:00:00 2015 -0500
  317.  
  318. library: add support for displaying LXC container name
  319.  
  320. This commit adds a lxc container name to every proc_t.
  321. If a process is not running in a container, then a '-'
  322. will be provided, making such a field always sortable.
  323.  
  324. Unlike other proc_t character pointers, lxc containers
  325. will find many duplicate shared values. So rather than
  326. strdup 'em (with a later free required upon reuse), we
  327. try to keep track of those already seen and share that
  328. address among all tasks running within each container.
  329.  
  330. We rely on the lines in the task's cgroup subdirectory
  331. which may initially seem somewhat unsophisticated. But
  332. the lxc library itself uses a similar approach when it
  333. is called to list active containers. In that case, the
  334. /proc/net/unix directory is parsed for the '/lxc' eye-
  335. catcher, with potential complications from hashed path
  336. and names that are too long (something we don't face).
  337.  
  338. [ too bad docker abandoned lxc - our commit won't do ]
  339. [ anything for the users of those kind of containers ]
  340.  
  341. Reference(s):
  342. https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/1424253
  343. https://bugs.launchpad.net/ubuntu/+source/procps/+bug/1424253
  344.  
  345. Signed-off-by: Jim Warner <james.warner@comcast.net>
  346.  
  347. commit 96bce4e11ee5bcbda8062169a75dea5561ef3618
  348. Author: Jim Warner <james.warner@comcast.net>
  349. Date: Sat Jun 13 00:00:00 2015 +0000
  350.  
  351. library: address memory aberration with LXC containers
  352.  
  353. Under a lxc container, the /proc/meminfo 'MemFree' and
  354. 'MemAvailable' amounts will be equal, unless memory is
  355. being limited via cgroups in which case 'MemAvailable'
  356. could exceed that for 'MemTotal'. And when a container
  357. has been nested, there exist additional memory quirks.
  358.  
  359. A program might then display used or available amounts
  360. greater than total memory (assuming unsigned honored),
  361. or negative values (should a signed cast be employed).
  362.  
  363. This anomaly primarily impacted the top and free pgms.
  364. Thus, two simple sanity checks have been introduced to
  365. avoid any illogical kb_main_available or kb_main_used.
  366.  
  367. ( Busybox top & free also display anomalous although )
  368. ( different results when running in a lxc container. )
  369.  
  370. Reference(s):
  371. https://bugzilla.redhat.com/show_bug.cgi?id=1153817
  372.  
  373. Signed-off-by: Jim Warner <james.warner@comcast.net>
  374.  
  375. commit e107f5d63bbc52f5181d8356b344e854a16a2c04
  376. Author: Jim Warner <james.warner@comcast.net>
  377. Date: Sun Feb 1 00:00:00 2015 -0600
  378.  
  379. top: miscellaneous accumulated changes to program code
  380.  
  381. This commit just tweaks top in the following respects:
  382.  
  383. . for alphabetic integrity, change 'INSP_hdr...' names
  384.  
  385. . eliminate the -Wsometimes-uninitialized warning that
  386. was found under OSX Yosemite (llvm 6.0/clang-600.0.56)
  387.  
  388. . update program 'comments' reflecting copyright dates
  389.  
  390. Signed-off-by: Jim Warner <james.warner@comcast.net>
  391.  
  392. commit 8bcdd2145d6a3ab805536108b9a3dda3bf568f15
  393. Author: Jim Warner <james.warner@comcast.net>
  394. Date: Sun Feb 1 00:00:00 2015 -0600
  395.  
  396. top: miscellaneous accumulated changes to man document
  397.  
  398. Signed-off-by: Jim Warner <james.warner@comcast.net>
  399.  
  400. commit da06b8fa5974ff8d651e45e55d52475c371054fe
  401. Author: Jim Warner <james.warner@comcast.net>
  402. Date: Tue Oct 28 00:00:00 2014 -0500
  403.  
  404. top: tweak forest view protections for forking anomaly
  405.  
  406. A recent commit eliminated the potential for a storage
  407. violation with forest view mode. It occurred when some
  408. program (erroneously?) created a lengthy forking loop.
  409. However, the associated commit message was misleading.
  410.  
  411. The message implied that an unexpected order following
  412. a sort on start_time was the cause of storage overruns
  413. and a 'char' used to track nesting level only distorts
  414. the display when it goes negative. Actually, the truth
  415. is really just the opposite. Any start_time sort quirk
  416. causes no harm while that 'char' can yield corruption.
  417.  
  418. Should some child end up sorted ahead of its parent by
  419. way of an extremely unlikely shared start_time the end
  420. result is such a child will be displayed unnested just
  421. like init or kthreadd along with all its own children.
  422.  
  423. However, if nesting levels exceeded 255 (and became 0)
  424. a massive array overrun could be triggered when such a
  425. task and *all* its children were added to an array for
  426. the second time. Exactly how much storage was violated
  427. depended on the number of children that zeroed process
  428. had spawned (hinted at via either SIGSEGV or SIGABRT).
  429.  
  430. The earlier commit limited nested levels to 100 so the
  431. root cause of the storage violation was already fixed.
  432. The potential for distorted nesting levels due to sort
  433. on start_time would seem to remain. But it's extremely
  434. unlikely that 2 tasks would share the same start_time.
  435.  
  436. Even so, a new #define has been introduced which makes
  437. top impervious to the order of tasks such that a qsort
  438. is no longer necessary (providing an init/systemd task
  439. exists & was harvested as the first task by readproc).
  440. It can be utilized if distorted nesting ever becomes a
  441. real issue. But since there is a 5-10% performance hit
  442. with that, we'll continue using start_time as default.
  443.  
  444. References(s):
  445. commit ce70017eb1927be51f73cbe0a0b4babcc502607e
  446.  
  447. Signed-off-by: Jim Warner <james.warner@comcast.net>
  448.  
  449. commit ce70017eb1927be51f73cbe0a0b4babcc502607e
  450. Author: Jim Warner <james.warner@comcast.net>
  451. Date: Thu Oct 23 00:00:00 2014 -0500
  452.  
  453. top: provide some protection against forking anomalies
  454.  
  455. This commit will eliminate a very nasty bug associated
  456. with top's forest view mode. It addresses a potential
  457. SIGSEGV/SIGABRT that was only encountered when another
  458. program (erroneously?) creates a lengthy forking loop.
  459.  
  460. If the growing list of nested children is sufficiently
  461. fast such that proc_t start_time is duplicated between
  462. children then the sort upon which top relies might not
  463. produce the expected order. That, in turn, could cause
  464. the forest_adds function to initially miss some child.
  465.  
  466. But that missed child would be caught by forest_create
  467. and eventually would cause our array boundary overrun.
  468. Such overrun occurs when some child of that originally
  469. *missed* child is found and a duplicate add attempted.
  470.  
  471. In correcting this bug we'll also use this opportunity
  472. to prohibit a borrowed proc_t padding byte (char) from
  473. going negative. If the nesting level exceeded 127, the
  474. effect was an "unnesting" with the snprintf width then
  475. viewed as flag+width also yielding left justification.
  476.  
  477. Henceforth, we'll limit nesting to 100 with subsequent
  478. children shown as " + ", not the usual " `- " prefix.
  479.  
  480. References(s):
  481. https://bugzilla.redhat.com/show_bug.cgi?id=1153642
  482. http://www.freelists.org/post/procps/Bug-in-the-forrest-view,6
  483.  
  484. Signed-off-by: Jim Warner <james.warner@comcast.net>
  485.  
  486. commit b0767bd391da2c868ccabe1f0213e1593f1e2eed
  487. Author: Jim Warner <james.warner@comcast.net>
  488. Date: Sun Sep 28 00:00:00 2014 -0500
  489.  
  490. top: ensure previously saved rcfile honored completely
  491.  
  492. When startup defaults were changed users with existing
  493. rcfiles would likely find their previous configuration
  494. was not being honored in all respects. The disparities
  495. involved Graphs modes and Summary/Task memory scaling.
  496.  
  497. This patch simply restores what was always intended as
  498. the proper behavior for previously saved config files.
  499.  
  500. References(s):
  501. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=762928
  502. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=762947
  503. . new startup defaults
  504. commit 8ef6cd91fc5df6372a93bc4a37cbad34ead9654e
  505.  
  506. Signed-off-by: Jim Warner <james.warner@comcast.net>
  507.  
  508. commit 8adf4acc033e25dfa68edbf4a8b7fc44148cd370
  509. Author: Jim Warner <james.warner@comcast.net>
  510. Date: Fri Aug 8 11:59:59 2014 -0500
  511.  
  512. top: final tweak to recent changes for new graph modes
  513.  
  514. Signed-off-by: Jim Warner <james.warner@comcast.net>
  515.  
  516. commit fd71717b9b760c988898f5eb0d1459f0b04cc739
  517. Author: Jim Warner <james.warner@comcast.net>
  518. Date: Thu Aug 7 00:00:00 2014 -0500
  519.  
  520. NEWS: add a few more for the release 3.3.10 change log
  521.  
  522. Signed-off-by: Jim Warner <james.warner@comcast.net>
  523.  
  524. commit e92b692932b8313f335009758b60f67b5adaadc7
  525. Author: Jim Warner <james.warner@comcast.net>
  526. Date: Thu Aug 7 00:00:00 2014 -0500
  527.  
  528. top: swat a potential buglet affecting new graph modes
  529.  
  530. This patch will cure a potential aberration associated
  531. with a terminal's size (SIGWINCH) and top's new graphs
  532. modes. The symptoms were a dangling tilde (~) plus the
  533. potential loss of a graph's right-most visual content.
  534.  
  535. The condition was only apparent when a %Cpu approached
  536. 100% usage. Also the apparent loss of content affected
  537. the 'block' graph only. With 'bar' graphs, that affect
  538. became the loss of proper right-most bar graph colors.
  539.  
  540. The cause was determined to be a combination of: 1) an
  541. unnecessary snprintf precision specification; and 2) a
  542. rounding quirk for any graphs which displayed distinct
  543. types of information (as for user/syst, used/unavail).
  544. These could then combine to produce an extra bar/block
  545. which, in turn, resulted in the truncation of a pseudo
  546. termcap attribute used by the show_special() function.
  547.  
  548. What was originally interpreted as an intractable race
  549. condition turns out to be just a self inflicted wound.
  550.  
  551. Reference(s):
  552. http://www.freelists.org/post/procps/top-Possible-bug-in-the-graphs,1
  553.  
  554. Signed-off-by: Jim Warner <james.warner@comcast.net>
  555.  
  556. commit bcbc3c5a029b8936c72063bf1175b42a97f184f2
  557. Author: Jim Warner <james.warner@comcast.net>
  558. Date: Wed Aug 6 00:00:00 2014 -0500
  559.  
  560. misc: result after checking all files for misspellings
  561.  
  562. Reference(s):
  563. https://github.com/lyda/misspell-check.git
  564.  
  565. Signed-off-by: Jim Warner <james.warner@comcast.net>
  566.  
  567. commit 5380ef902297ad93b4691ed321cb85a727591505
  568. Author: Jim Warner <james.warner@comcast.net>
  569. Date: Wed Jul 30 00:00:00 2014 -0500
  570.  
  571. top: adapt global memory support to new library format
  572.  
  573. Signed-off-by: Jim Warner <james.warner@comcast.net>
  574.  
  575. commit c437faf8d32c7cf5d2b0eefb7dc935a9778e5413
  576. Author: Jim Warner <james.warner@comcast.net>
  577. Date: Sat Jul 19 04:44:44 2014 -0500
  578.  
  579. library: evolve MenAvailable algorithm on older kernel
  580.  
  581. Let's not report zero for kb_main_available when older
  582. kernels don't have MemAvailable. Instead, if we simply
  583. duplicate the 'free' amount we can avoid all ancillary
  584. problems, such as those involving top's graphing mode.
  585.  
  586. Reference(s):
  587. http://www.freelists.org/post/procps/kb-main-available-etc,3
  588.  
  589. Signed-off-by: Jim Warner <james.warner@comcast.net>
  590.  
  591. commit 7391699b0748724b5d3024df4f01d2571bd29fc1
  592. Author: Jim Warner <james.warner@comcast.net>
  593. Date: Fri Jul 18 08:55:08 2014 -0500
  594.  
  595. top: correct one grammatical error in the man document
  596.  
  597. Signed-off-by: Jim Warner <james.warner@comcast.net>
  598.  
  599. commit d310a18fc2391ba06f3531aff020c97057c0ed3c
  600. Author: Jim Warner <james.warner@comcast.net>
  601. Date: Thu Jul 17 13:13:13 2014 -0500
  602.  
  603. top: exploit new kb_main_available, make Jaromir happy
  604.  
  605. This patch will trade a former pessimistic calculation
  606. of free physical memory for a more optimistic one that
  607. uses the newly added kb_main_available library export.
  608.  
  609. But in case one might wish to return to the old former
  610. method, there's a new #define that was made available.
  611.  
  612. [ the new calculation will affect graphing mode only ]
  613.  
  614. Reference(s):
  615. http://www.freelists.org/post/procps/systemd-support-to-library,9
  616.  
  617. Signed-off-by: Jim Warner <james.warner@comcast.net>
  618.  
  619. commit 9500dc198c6347768a71cdf4a6a2a8d5cba903f3
  620. Author: Jim Warner <james.warner@comcast.net>
  621. Date: Thu Jul 17 12:12:12 2014 -0500
  622.  
  623. top: trade Page_size for that newly exposed page_bytes
  624.  
  625. Might as well use the newly exposed sysinfo.h variable
  626. 'page_bytes' rather than our own. Plus, in the process
  627. we can avoid incurring yet one more function call hit.
  628.  
  629. Signed-off-by: Jim Warner <james.warner@comcast.net>
  630.  
  631. commit e4f0b4ebf3eaa1bda9a0315ec7b113c73a510dd4
  632. Author: Jim Warner <james.warner@comcast.net>
  633. Date: Thu Jul 17 11:11:11 2014 -0500
  634.  
  635. library: tweak newly added MemAvailable fall-back code
  636.  
  637. This commit just ensures recalculation of some amounts
  638. for iterative processes, like top. It also trades some
  639. repeated runtime calls to sysconf for a one time cost.
  640.  
  641. Reference(s):
  642. http://www.freelists.org/post/procps/systemd-support-to-library,7
  643. . fall-back calculations
  644. commit b779855cf15d68f9038ff1809db18c0788e9ae70
  645.  
  646. Signed-off-by: Jim Warner <james.warner@comcast.net>
  647.  
  648. commit 6c148b4f8b2acfaf1edae9f8e92e91bdeb723e72
  649. Author: Jim Warner <james.warner@comcast.net>
  650. Date: Thu Jul 17 10:10:10 2014 -0500
  651.  
  652. docs: with systemd now in library, remember openproc.3
  653.  
  654. [ plus remove just a little darn trailing whitespace ]
  655.  
  656. Reference(s):
  657. . systemd migrated to library
  658. commit 9d8ad6419f9f16978b7e9e769b7acb085f317058
  659. . added library documentation
  660. commit a74fb8fade82ad624703bb1bc7395cf0aacaeaa9
  661.  
  662. Signed-off-by: Jim Warner <james.warner@comcast.net>
  663.  
  664. commit 1da2c98937d91676e485557f8197249453547306
  665. Author: Jim Warner <james.warner@comcast.net>
  666. Date: Wed Jul 16 00:00:00 2014 -0500
  667.  
  668. library: disable a potential 'ELF note' is missing msg
  669.  
  670. The stderr message regarding ELF notes appears on some
  671. systems (openSUSE-13.1 for example) but I have not yet
  672. isolated why. Since at startup we go on to determine a
  673. Hertz value the old fashion way, this patch just turns
  674. off the useless message until the cause is understood.
  675.  
  676. Signed-off-by: Jim Warner <james.warner@comcast.net>
  677.  
  678. commit 93d37cf57a9d3c4437c5ed660007d2c471c6daee
  679. Author: Jim Warner <james.warner@comcast.net>
  680. Date: Sat Jun 28 00:00:33 2014 -0500
  681.  
  682. ps: exploit library systemd support vs. internal logic
  683.  
  684. Signed-off-by: Jim Warner <james.warner@comcast.net>
  685.  
  686. commit 9d8ad6419f9f16978b7e9e769b7acb085f317058
  687. Author: Jim Warner <james.warner@comcast.net>
  688. Date: Sat Jun 28 00:00:22 2014 -0500
  689.  
  690. library: generalize then add former 'ps' systemd logic
  691.  
  692. Signed-off-by: Jim Warner <james.warner@comcast.net>
  693.  
  694. commit d18730485435188946ca773c372cca162a9ad696
  695. Author: Jim Warner <james.warner@comcast.net>
  696. Date: Sat Jun 28 00:00:11 2014 -0500
  697.  
  698. ps: clarify which options depend on systemd in man doc
  699.  
  700. Signed-off-by: Jim Warner <james.warner@comcast.net>
  701.  
  702. commit 6cd8691720a470c20b5aace7186a7f1984d82930
  703. Author: Jim Warner <james.warner@comcast.net>
  704. Date: Fri Jul 4 00:00:00 2014 -0500
  705.  
  706. top: fix potential 'nan', should a system have no Swap
  707.  
  708. Gosh, just because most of us might run with some swap
  709. file allocated, not every system might. I only wish my
  710. testing methodology was as sophisticated as Jaromir's.
  711.  
  712. Reference(s):
  713. http://www.freelists.org/post/procps/latest-top-enhancements,7
  714.  
  715. Signed-off-by: Jim Warner <james.warner@comcast.net>
  716.  
  717. commit 1dd0c4d07f870351dedfbee83da0a140f76292b2
  718. Author: Jim Warner <james.warner@comcast.net>
  719. Date: Thu Jul 3 00:00:00 2014 -0500
  720.  
  721. top: as with ps, distinguish between 'T' and 't' state
  722.  
  723. Wow, even the linux/Documentation/filesystems/proc.txt
  724. document doesn't provide us with that level of detail.
  725.  
  726. Thank you, Jaromir, for your additional clarification!
  727.  
  728. Reference(s):
  729. commit 411d218793af1f7ef6db37b319c1b08a7ca7c9b8
  730.  
  731. Signed-off-by: Jim Warner <james.warner@comcast.net>
  732.  
  733. commit a7741055f19c4273c7c17c8af4f89195d018c2ba
  734. Author: Jim Warner <james.warner@comcast.net>
  735. Date: Wed Jul 2 00:00:00 2014 -0500
  736.  
  737. top: add another translation hint for graphs alignment
  738.  
  739. This should be the last of this kind of crap. I'll get
  740. to work on some means to no longer burden a translator
  741. with lengths requirements. Ideally each word should be
  742. allowed to stand alone and the minimum/maximum lengths
  743. handled programmatically when our ol' top is executed.
  744.  
  745. Signed-off-by: Jim Warner <james.warner@comcast.net>
  746.  
  747. commit 206570e8c453104457b6179ff922e74d3a5fba96
  748. Author: Jim Warner <james.warner@comcast.net>
  749. Date: Tue Jul 1 00:00:00 2014 -0500
  750.  
  751. top: with new startup defaults, tweak the man document
  752.  
  753. Signed-off-by: Jim Warner <james.warner@comcast.net>
  754.  
  755. commit c75586f5238fb65708404641eb6cde6c1a403c84
  756. Author: Jim Warner <james.warner@comcast.net>
  757. Date: Mon Jun 30 00:00:00 2014 -0500
  758.  
  759. top: eliminated unreferenced macros & an error message
  760.  
  761. Signed-off-by: Jim Warner <james.warner@comcast.net>
  762.  
  763. commit 8ef6cd91fc5df6372a93bc4a37cbad34ead9654e
  764. Author: Jim Warner <james.warner@comcast.net>
  765. Date: Sun Jun 29 00:00:00 2014 -0500
  766.  
  767. top: retire old stale startup defaults in favor of new
  768.  
  769. For over a decade top has used a startup configuration
  770. mimicking the original redhat top. This decision dates
  771. back to when the forked Sourceforge version was trying
  772. to win over users in battles with that ancient kludge.
  773.  
  774. Will anybody deny that those defaults are coyote ugly?
  775.  
  776. Well, it is time that top presented a more modern look
  777. at startup, providing that no saved rcfile exists. But
  778. just in case some distro prefers that old, comfortable
  779. look, there's the '--disable-modern-top' build option.
  780.  
  781. [ Pssst. With the widened memory fields it turns out ]
  782. [ the 'Mem' default window had become almost useless ]
  783. [ on an 80x24 terminal since %CPU & COMMAND were out ]
  784. [ of view. So some other defaults were tweaked a bit ]
  785. [ whether or not --disable-modern-top was specified. ]
  786.  
  787. Reference(s)
  788. http://www.freelists.org/post/procps/tops-graph-mode-saga-continues,3
  789.  
  790. Signed-off-by: Jim Warner <james.warner@comcast.net>
  791.  
  792. commit 2199af404a92cac238728564f2f022255c5716ce
  793. Author: Jim Warner <james.warner@comcast.net>
  794. Date: Sat Jun 28 00:00:00 2014 -0500
  795.  
  796. top: maximize recent locale aware numeric enhancements
  797.  
  798. When startup argument parsing was recently enhanced to
  799. account for LC_NUMERIC settings, some user input logic
  800. dealing with numbers fails to exploit that capability.
  801.  
  802. This patch extends such enhancements to a running top.
  803.  
  804. Reference(s):
  805. commit f7b84f45c7ae99c276de9954fc16cdc4ff7f36f0
  806. http://www.freelists.org/post/procps/topwatch-floating-point-input,2
  807.  
  808. Signed-off-by: Jim Warner <james.warner@comcast.net>
  809.  
  810. commit 96c330e3b33aeec75d5210b8bd9e4c33e4135a13
  811. Author: Jim Warner <james.warner@comcast.net>
  812. Date: Fri Jun 27 00:00:00 2014 -0500
  813.  
  814. top: afford each window its own cpu/memory graph modes
  815.  
  816. When those new cpu/memory graphs modes were introduced
  817. they had global impact. In other words, the modes that
  818. were chosen for a 'current' window affect Summary Area
  819. appearance for every other window as well, even though
  820. each window sets unique View_STATES/View_MEMORY flags.
  821.  
  822. I do not know how widespread the use of top's separate
  823. window provisions is, but I do know that documentation
  824. promises every window (field group) provides "a unique
  825. separately configurable summary area". And even though
  826. that promise does not include memory scaling (separate
  827. 'E' command) the graph modes are integral to 't' & 'm'
  828. and those were already observed on a per window basis.
  829.  
  830. So this patch just takes the cpu and memory graph mode
  831. values out of global scope in the configuration file &
  832. gives each window its own unique pair of graph values.
  833.  
  834. Reference(s):
  835. commit 1d171ec74163f3103e7b230ee63818df04b4526f
  836.  
  837. Signed-off-by: Jim Warner <james.warner@comcast.net>
  838.  
  839. commit 805532ac32610279f381e2510b8cbaac6c625903
  840. Author: Jim Warner <james.warner@comcast.net>
  841. Date: Thu Jun 26 00:00:00 2014 -0500
  842.  
  843. top: fix translation hint to preserve graphs alignment
  844.  
  845. Signed-off-by: Jim Warner <james.warner@comcast.net>
  846.  
  847. commit f7b84f45c7ae99c276de9954fc16cdc4ff7f36f0
  848. Author: Jim Warner <james.warner@comcast.net>
  849. Date: Wed Jun 25 00:00:00 2014 -0500
  850.  
  851. top: tweak argument parsing for some locale situations
  852.  
  853. Boy I hate locale stuff. For code I thought was pretty
  854. robust, Jaromir sure proved that it wasn't. Anyway, me
  855. thinks this commit closes some gaps and will cause top
  856. to behave appropriately under various locale settings.
  857.  
  858. It does *not* permit top to respond to the ',' and '.'
  859. floating point separator without regard to the locale.
  860. It does, however, enforce proper LC_NUMERIC responses.
  861.  
  862. Let's look on this commit as an interim solution until
  863. Jaromir can create that proposed 'fp_decode' function.
  864. Who knows, he might even borrow some of our mkfloat().
  865.  
  866. [ An aside: the coreutils sleep and timeout programs ]
  867. [ claim to permit floating point arguments. However, ]
  868. [ neither one will accept the comma separator should ]
  869. [ the locale be a country that in fact uses a comma. ]
  870.  
  871. [ In other words, with this commit we are way ahead! ]
  872.  
  873. Reference(s):
  874. http://www.freelists.org/post/procps/topwatch-floating-point-input
  875. http://www.freelists.org/post/procps/topwatch-floating-point-input,1
  876.  
  877. Signed-off-by: Jim Warner <james.warner@comcast.net>
  878.  
  879. commit cec19765115e9c65c52412c16d03f8d515506d7e
  880. Author: Jim Warner <james.warner@comcast.net>
  881. Date: Tue Jun 24 00:00:00 2014 -0500
  882.  
  883. top: let's not pretend top can catch SIGKILL & SIGSTOP
  884.  
  885. While there was no harm done setting a handler for the
  886. above two signals, they are in fact uncatchable. Thus,
  887. whenever we ran with valgrind we're politely reminded.
  888.  
  889. Signed-off-by: Jim Warner <james.warner@comcast.net>
  890.  
  891. commit db6381ae043ffd4aa9fd5788f434af155ae096f1
  892. Author: Jim Warner <james.warner@comcast.net>
  893. Date: Mon Jun 23 00:00:00 2014 -0500
  894.  
  895. top: scale length for new graphs to a terminal's width
  896.  
  897. When the beginning of the Mem/Swap graphs was variable
  898. scaling them to the current terminal's width was a bit
  899. of a costly nightmare. So the graph size was fixed and
  900. subject to truncation. However now that the start of a
  901. graph can be easily predicted, I've revisited scaling.
  902.  
  903. As it turns out, any cost is minimal & mostly incurred
  904. at an opportune time, at SIGWINCH or user interaction.
  905. Plus, most of the apparent arithmetic is actually just
  906. a means of documenting and will disappear thru compile
  907. time constants in the ultimate generated machine code.
  908.  
  909. Note: those graphs will now behave just like any other
  910. Summary Area element - they will scale from full sized
  911. down to a terminal width of 80 columns, at which point
  912. those displayed graphs are then subject to truncation.
  913.  
  914. Signed-off-by: Jim Warner <james.warner@comcast.net>
  915. ...
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!