Description: With the steps specified below, cal fails to properly recalculate. The demo sheet will be provided, but here are the steps to create it: - column a: sequential numbers from 1 to 9, but swap 4 and 5 (1,2,3,5,4,6,7,8,9) - b1: 1 - c1: =sum(b$1:b1) - copy b1:c1 to b2:c9 Steps to Reproduce: 1. Select rows 3 to 6 2. Data -> Sort ascending. The sort completes as expected 3. Change b4 from 1 to 10 Actual Results: c1:c6 contains correct values. c7:c9 don't Forcing a recalculate (F9) don't resolve the situation. Forcing a hard recalculate (ctrl-shift-F9) resolves it. Expected Results: Correct auto-recalculation of all cells. Reproducible: Always User Profile Reset: No Additional Info: Initially observed on the version provided by linux/debian. Formally tested on Windows 10 in safe mode, after deleting the LibreOffice folder in appdata/roaming. No idea if OpenGL is enabled. Info from help-about: Version: 7.4.0.3 (x64) / LibreOffice Community Build ID: f85e47c08ddd19c015c0114a68350214f7066f5a CPU threads: 4; OS: Windows 10.0 Build 19044; UI render: default; VCL: win Locale: en-US (en_US); UI: en-US Calc: threaded
Created attachment 182186 [details] bug example
I happens also with Version: 7.5.0.0.alpha0+ (x64) / LibreOffice Community Build ID: 21229c2f45f96a8ee3060acb3296be77be3a9f1a CPU threads: 4; OS: Windows 10.0 Build 19044; UI render: Skia/Vulkan; VCL: win Locale: es-ES (es_ES); UI: en-US Calc: CL threaded
Repro Version: 7.1.8.0.0+ (x64) / LibreOffice Community Build ID: a94b58277c7aeaa83ce14347cd0b8f7137969d03 CPU threads: 4; OS: Windows 6.3 Build 9600; UI render: Skia/Raster; VCL: win Locale: nl-NL (nl_NL); UI: en-US Calc: CL and with Version: 6.2.9.0.0+ (x86) Build ID: 5f01fe15eb2661f1f9ce12d1d99dc2a705b462ee CPU threads: 4; OS: Windows 6.3; UI render: default; VCL: win; Locale: nl-NL (nl_NL); UI-Language: en-US Calc: CL and with Version: 6.0.0.0.alpha0+ Build ID: 9127d1a89cbfba89eb9df6755ea7b9e161cfc67a CPU threads: 4; OS: Windows 6.3; UI render: default; Locale: nl-NL (nl_NL); Calc: CL and with Versie: 4.4.7.2 Build ID: f3153a8b245191196a4b6b9abd1d0da16eead600 Locale: nl_NL ok with Versie: 4.1.0.4 Build ID: 89ea49ddacd9aa532507cbf852f2bb22b1ace28
@Eike, I thought you might be interested in this issue.
It started to happen in libreoffice-4.4.0.0.beta1, thus it needs to be bisected with 4.4 repo
d12d1a450eb7599eb60ba605ce7a7e8b12bcb744 is the first bad commit commit d12d1a450eb7599eb60ba605ce7a7e8b12bcb744 Author: Matthew Francis <mjay.francis@gmail.com> Date: Sun Mar 15 06:07:23 2015 +0800 source-hash-5ef856e975d7c0396984d588a43fd1a7c7085c55 (Bibisect: Skipped preceding irrelevant commit(s) b2efe90c07baa45d2ee7e13138dd306a29738473) commit 5ef856e975d7c0396984d588a43fd1a7c7085c55 Author: Kohei Yoshida <kohei.yoshida@collabora.com> AuthorDate: Tue Nov 18 21:03:05 2014 -0500 Commit: Kohei Yoshida <kohei.yoshida@collabora.com> CommitDate: Tue Nov 18 21:03:50 2014 -0500 Use group area listeners during ODS import. Change-Id: Id01f9021dda7f33255f8206174cd730507ab55ad :040000 040000 126e010c62ce770ece406648bd72b0d2505c593f e7bee6d7439acbe7be3fd506baefed81c493b191 M opt Bisected using https://bibisect.libreoffice.org/linux-44max.git-bundle The bisect log isn't "normal" - the very first bisection step was unconclusive, LibreOffice consistently crashed on the sort step, so I kickstarted it testing a random point. # bad: [cf6ea17155fabb2a120ba07c150735591ac861d7] source-hash-3f94c9e9ddfd807b449f3bb9b232cf2041fa12d2 # good: [fc71ac001f16209654d15ef8c1c4018aa55769f5] source-hash-c15927f20d4727c3b8de68497b6949e72f9e6e9e git bisect start 'latest' 'oldest' # skip: [8cf60cc706948588e2f33a6d98b7c55d454e362a] source-hash-f340f0454627939f1830826fb5cc53a90e6c62a4 git bisect skip 8cf60cc706948588e2f33a6d98b7c55d454e362a # good: [2d15a9a7574c794c9dc50f5ff4cd5ed4464fde70] source-hash-a6fc653bbf38de453a540ca2ab69958dafa1d322 git bisect good 2d15a9a7574c794c9dc50f5ff4cd5ed4464fde70 # good: [9a82af5ef45cf613dfab0f6e979d404b717018f9] source-hash-96d091d43428472c2c68054ed5dca22f79adc705 git bisect good 9a82af5ef45cf613dfab0f6e979d404b717018f9 # good: [4f7b9e3948b9f7b42c81f5c1293aa73491001c93] source-hash-fa430e6b4e6f5d096bdf59db26e5d7393ca2297b git bisect good 4f7b9e3948b9f7b42c81f5c1293aa73491001c93 # good: [0fbc26d3c8e7c48fb6d13aa99006f355601b5f7d] source-hash-9313095ea2d5f3dcc531d658e159e16ac2cb44fd git bisect good 0fbc26d3c8e7c48fb6d13aa99006f355601b5f7d # good: [8fbd6faad6cc859f6f2ab984a031b818d9471140] source-hash-65f41639e219470c80f896129731f9ed6f072b6b git bisect good 8fbd6faad6cc859f6f2ab984a031b818d9471140 # good: [e076716822b668af0bd9f480a6f67336e0d1591b] source-hash-dbdacc73ae154237314d599194cc686bd738a9bf git bisect good e076716822b668af0bd9f480a6f67336e0d1591b # bad: [40de7a9096dba2cd21825772dc4fd776c6f7fdaa] source-hash-78ac39dcef36b3881b2574b2db1e0d5a068c7804 git bisect bad 40de7a9096dba2cd21825772dc4fd776c6f7fdaa # bad: [57ed981e95cf3b915ed02bd1f8d2a9ecf618351c] source-hash-f857358d83e7c105271eb0e2c43f0b036f14f284 git bisect bad 57ed981e95cf3b915ed02bd1f8d2a9ecf618351c # good: [8a43f1c48415bd9e44ad81ca13c31ca39d3b15e9] source-hash-a55adb16ece70360c88342ca008d5a9d5b9d5b52 git bisect good 8a43f1c48415bd9e44ad81ca13c31ca39d3b15e9 # bad: [7c712fe89744ed1190bcee076bdbf81d40f1dc85] source-hash-3f4896bcfe21794239f8ec32f8f6d7e438a7907e git bisect bad 7c712fe89744ed1190bcee076bdbf81d40f1dc85 # good: [ad806e9b6fa58c2589859cb3c55511235f2c534f] source-hash-1296f9063203a3f9488689136b833568c183edab git bisect good ad806e9b6fa58c2589859cb3c55511235f2c534f # bad: [52dc3a6032f416dfe4a7182adc1d187ffcd9af24] source-hash-764600bc6175a32d119487afda2e8409f3395666 git bisect bad 52dc3a6032f416dfe4a7182adc1d187ffcd9af24 # bad: [6ca58fb62944a6ef674e4630740673971334f424] source-hash-1b9aaba0bfe8bc0872e7ea9f9aef5961e4b52f7c git bisect bad 6ca58fb62944a6ef674e4630740673971334f424 # bad: [d12d1a450eb7599eb60ba605ce7a7e8b12bcb744] source-hash-5ef856e975d7c0396984d588a43fd1a7c7085c55 git bisect bad d12d1a450eb7599eb60ba605ce7a7e8b12bcb744 # first bad commit: [d12d1a450eb7599eb60ba605ce7a7e8b12bcb744] source-hash-5ef856e975d7c0396984d588a43fd1a7c7085c55
Thanks for bibisect, it's correct, some remarks: 1. instructions for bibisect advise to paste in bug the full log, but I don't see it necessary 2. more important is to check result with found and previous binary commit 3. what you get in older repos is not a single result, but a range of source commits from previous to found binary commit Here, it's just 2 source commits so that one is probably correct (so I add bisected). https://cgit.freedesktop.org/libreoffice/core/log/?qt=range&q=1296f9063203a3f9488689136b833568c183edab..5ef856e975d7c0396984d588a43fd1a7c7085c55 4. then we check is source commit is already in some comments, this one is but all are bugs fixed by Kohei and mostly Eike: bug 86615, bug 97897, bug 121043, bug 119623. 5. finally may check if dev is active: https://cgit.freedesktop.org/libreoffice/core/log/&qt=author&q=Kohei+Yoshida Kohei is active sometimes, so I add him here to see this. I hope that kohei@libreoffice.org is the right address, there's also another one.
Taking.
Eike Rathke committed a patch related to this issue. It has been pushed to "master": https://git.libreoffice.org/core/commit/569d451bbe160829947ace201d07ad69442204b0 Resolves: tdf#150766 Listen to entire shared formula group instead of a slice It will be available in 7.5.0. The patch should be included in the daily builds available at https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: https://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.
Pending review https://gerrit.libreoffice.org/c/core/+/139662 for 7-4 https://gerrit.libreoffice.org/c/core/+/139663 for 7-3
Will there also be a unit test for this bug?
@b. FYI
Verified in Version: 7.5.0.0.alpha0+ / LibreOffice Community Build ID: 9caf9f8fde68f075a9ae1377bcc0cf6127c1737f CPU threads: 8; OS: Linux 5.10; UI render: default; VCL: gtk3 Locale: es-ES (es_ES.UTF-8); UI: en-US Calc: threaded @Eike, thanks for fixing this issue!!
Eike Rathke committed a patch related to this issue. It has been pushed to "libreoffice-7-3": https://git.libreoffice.org/core/commit/dd51e61b9fcdb0833b89be3aa7dd57a3d9e00c7b Resolves: tdf#150766 Listen to entire shared formula group instead of a slice It will be available in 7.3.7. The patch should be included in the daily builds available at https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: https://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.
Eike Rathke committed a patch related to this issue. It has been pushed to "libreoffice-7-4": https://git.libreoffice.org/core/commit/23d607b104013e412295ac83956edc74e94ed2df Resolves: tdf#150766 Listen to entire shared formula group instead of a slice It will be available in 7.4.2. The patch should be included in the daily builds available at https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: https://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.
Xisco Fauli committed a patch related to this issue. It has been pushed to "master": https://git.libreoffice.org/core/commit/907ba54b0d557b89e7e0e6f6eb25470e743c1918 tdf#150766: sc_uicalc: Add unittest It will be available in 7.5.0. The patch should be included in the daily builds available at https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: https://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.