summaryrefslogtreecommitdiff
path: root/rsc
diff options
context:
space:
mode:
authorStephan Bergmann <sbergman@redhat.com>2017-02-21 07:37:08 +0100
committerStephan Bergmann <sbergman@redhat.com>2017-02-21 07:37:08 +0100
commite972d6bb36efaa14eb4983c835491cbcd6283e91 (patch)
treea7ca6a03fd12921860ab630ef4a10cc89b7e4200 /rsc
parent4c8154bfd97fe650d2222dc095b18150a10dc626 (diff)
Idlc::m_pErrorHandler is unused
Change-Id: I41fea5e08381e21dff12f61142a4d54f07a502f4
Diffstat (limited to 'rsc')
0 files changed, 0 insertions, 0 deletions
generated by cgit (git 2.34.1) at 2025-01-21 17:02:08 +0000
e it's guaranteed. If we raise the baseline to AVX, the SSE2 code may be replaced by the one removed by this commit. Generic code is there for other platforms, if other platforms add CPU-specific code, they should preferably follow the same rules. This does not necessarily mean that CPU-specific code cannot be used at all. Some externals use them, for example. It just needs to be working, maintained, and worth the trouble. Change-Id: I5ab919930df9d0223db68a94bf84947984d313ac Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129733 Tested-by: Jenkins Reviewed-by: Eike Rathke <erack@redhat.com> 2022-01-15tdf#146367 reintroduce approxAdd() similar handling with last summandEike Rathke ... to tie result to 0 eliminating opposite sign small equalish fractions. This does not "fix" all interim values of the bug's sample document, which repeatedly adds single values in sequence using SUM() and should use the + operator instead to benefit from approxAdd() for each, but maintains the better accuracy of Kahan summation in the not near equal zero cases. And the dreaded SUM(0.1;0.2;-0.3) also works again resulting in 0 with all permutations. Change-Id: I67188a6abbdb98212f070166ad02319c3d477e28 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128451 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins 2021-09-13properly separate code built with different CPU settingsLuboš Luňák Trying to write smart code and mixing different CPU flags doesn't play nice together. Those global variables are not runtime-protected by a CPU check, and so may crash with illegal instruction error. And those inline functions may not get inlined and the compiler is free to choose just one copy, any of them, so it may be the one requiring the most demanding CPU settings. So use only dumb code in files compiled with CPU intrinsics. Change-Id: I8200fd4d9f991fab6fdc741120e7aa96ff9b470d Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121929 Tested-by: Jenkins Reviewed-by: Dante DM <dante19031999@gmail.com> Reviewed-by: Luboš Luňák <l.lunak@collabora.com> 2021-08-26tdf#142307 - Upgrade SSE2 sum to AVX512 sum with Neumaier 1dante This part focuses on allowing it on replacing arrayfunctor By thefault it will try AVX512F (1,17%) If not available will use AVX (94,77%) Use of AVX2 (82,28%) has been avoided even if the code could been more compact Source of hardware statistics: https://store.steampowered.com/hwsurvey Change-Id: Iae737a565379e82c5f84f3fdee6321ac74f59d40 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115675 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> 2021-05-08Simplify multiplication and divisionMike Kaganski Use operators that take double for unification Change-Id: I5ef6f8a684f35dffbd639435415547d6dc96ed58 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115208 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> 2021-05-08tdf#137679 Use Kahan summation for interpr3.cxx (2/2)dante Change-Id: I64113449f70d300feace6663c670db3448f43acf Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114970 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> 2021-05-04tdf#137679 Use kahan summation for ScInterpreter::lcl_IterateInversedante Change-Id: I02e108ac70ddd4ea8d8d97eb4f5fbc8996dd4bd1 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114966 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> 2021-04-29tdf#137679 Use Kahan summation for interpr2.cxxdante The changes in the text files are in the latest decimals. Those should be more accurate now. Change-Id: I3814e1939f71debd5ddde9408a025af7a0a2cac5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114737 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> 2021-04-28Fix typosAndrea Gelmini Change-Id: Ic73ee21dea2bc99c5fd8af857a9793782d7c3301 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114747 Tested-by: Jenkins Reviewed-by: Hossein <hossein.ir@gmail.com> Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com> 2021-04-26Add a Kahan summation class.dante Change-Id: I4021d07a2c011c2d5b2e79ecd878019d29cabd21 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114567 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>