VistaLover Posted April 29 Posted April 29 (edited) 6 hours ago, roytam1 said: upstream tested and it needs to be reverted. https://forum.palemoon.org/viewtopic.php?p=272957#p272957 ... Not reverted per se in the end , but modified to cater for old single or double core CPUs : https://repo.palemoon.org/MoonchildProductions/UXP/issues/3066 https://repo.palemoon.org/MoonchildProductions/UXP/commit/3259bdc1c14deea4c4ec3133901b783ea37d5c09 Edited April 29 by VistaLover 1
roytam1 Posted April 29 Author Posted April 29 (edited) 10 hours ago, VistaLover said: ... Not reverted per se in the end , but modified to cater for old single or double core CPUs : https://repo.palemoon.org/MoonchildProductions/UXP/issues/3066 https://repo.palemoon.org/MoonchildProductions/UXP/commit/3259bdc1c14deea4c4ec3133901b783ea37d5c09 but it is still can be overloading CPU as threadCount = cpuCount+4: https://github.com/roytam1/UXP/blob/69d0209956c29ea5ec21eaaa6414773389445d6d/js/src/vm/HelperThreads.cpp#L77 BTW I wonder if LoongArch changes can be imported to moebius repo without big changes. EDIT: decided not to merge LoongArch changes. Edited April 30 by roytam1 1
VistaLover Posted April 29 Posted April 29 (edited) 8 hours ago, roytam1 said: I was thinking if it is caused by https://github.com/roytam1/UXP/commit/ae7c40d4611542893882b2a3ea5d739852962734 Thanks for identifying this ; so, a backport from "dactyloidae" ; well, as we all know, NM28 doesn't support WE at all, so, IMHO, it should've been excluded from this change, though I realise that the change is platform-wide, not app-specific... In any case, that commit (that upstream won't ever merge, as they don't support WE at all) was probably NOT accurately authored, because it disables the standard installation of a category (with strictCompatibility=true) of non-WE (i.e. XUL/legacy) addons that the platform itself (UXP) should support natively/out-of-the-box; in addition, have you actually bestowed upon Serpent 52 WE-support on par with fx-128? I am but a casual UXP user by now , but this looks like a true regression to me ... What are your feelings about this? Are you inclined to fix or revert that commit? Or should one manually modify the XPIs of ALL addons with strictCompatibility=true to have them installed/re-enabled in UXP apps after the 20260418 updates? Thanks for your precious time! Edited April 29 by VistaLover better English syntax 2
Leokids123 Posted April 29 Posted April 29 Whoever made these changes should have been reverted due to what it resulted into...
roytam1 Posted April 30 Author Posted April 30 9 hours ago, VistaLover said: Thanks for identifying this ; so, a backport from "dactyloidae" ; well, as we all know, NM28 doesn't support WE at all, so, IMHO, it should've been excluded from this change, though I realise that the change is platform-wide, not app-specific... In any case, that commit (that upstream won't ever merge, as they don't support WE at all) was probably NOT accurately authored, because it disables the standard installation of a category (with strictCompatibility=true) of non-WE (i.e. XUL/legacy) addons that the platform itself (UXP) should support natively/out-of-the-box; in addition, have you actually bestowed upon Serpent 52 WE-support on par with fx-128? I am but a casual UXP user by now , but this looks like a true regression to me ... What are your feelings about this? Are you inclined to fix or revert that commit? Or should one manually modify the XPIs of ALL addons with strictCompatibility=true to have them installed/re-enabled in UXP apps after the 20260418 updates? Thanks for your precious time! alright, that rev is reverted. 1
johk Posted April 30 Posted April 30 (edited) So, in the end, I was right with my early findings about the CPU usage even Moonchild had to confirm I just really hope that those thread parsing corrections help to make NewMoon/Goanna engine smoother, on CPU I mean. The actual smooth would be accomplished on the javascript engine (as Firefox 115), someday... ------- FOLLOWS, AFTER ALL OFF-TOPIC ----- Sorry if I haven't replied these days, I've been testing RAM (that you know it takes loooooooot of time) with thunderstorms in the middle that didn't allow me to even plug the PC and now I'm fixing problems on a family member PC. And the RAM and processor are fine. In the meantime I rechecked all my previous debuglogs (made with debug wizard help) and, even the "probably caused by: win32k.sys" is very... (as any other with microsoft) imprecise, and I never considered it to be the real problem, I searched and applied the latest(?) win32k.sys update named KB3095649 which, in theory, might help, but as told, I haven't had time to test yet, because I was able to reproduce intentionally the BSOD, so I guess I could test. Answering to other mentions, the disk is not the problem as, long ago, decided to run the profile on a ramdisk (made with ImDisk) thanks to the decision long ago form Mozilla, and so MoonChild, and so Gecko and Goanna engine, to use an sqlite database for the history. You can't believe the performance hit that is managing those databases on disk, really. It is a continous read/write. And Firefox, using localStorage on sqlite databases for browser cache... you couldn't believe the performance difference of running it on a ramdisk. Anyway, I'm digressing right? I will test if NewMoon now crashes the system (not because of NewMoon but because I found a good testing vector), but that is something on my side. EDIT 2026-05-01T03:30+2: for what I tested before go to sleep, I couldn't reproduce the BSOD, so, fingers crossed, let's hope I don't have new BSODs with NewMoon as vector for win32k.sys BSOD, nor any other program. Or maybe today wasn't the same environment and didn't crash. The upcoming days will tell. And I hope the newer NewMoon builds, without the Goanna engine CPU hog, would be helpful as well. Edited May 1 by johk last edit information about my bsods "related" to newmoon 2
j7n Posted May 1 Posted May 1 What is the exact stop error that you get on the BSOD? I got CSRSS ("important process") terminating, and before that there was a lag and programs displaying memory read errors.
roytam1 Posted May 1 Author Posted May 1 New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20260502-3219d2d-uxp-9161cd3bdb-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20260502-3219d2d-uxp-9161cd3bdb-xpmod.7z source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/custom IA32 Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20260502-3219d2d-uxp-9161cd3bdb-xpmod-ia32.7z source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/ia32 NM28XP build: Win32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260502-d849524bd-uxp-9161cd3bdb-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260502-d849524bd-uxp-9161cd3bdb-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260502-d849524bd-uxp-9161cd3bdb-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260502-d849524bd-uxp-9161cd3bdb-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260502-d849524bd-uxp-9161cd3bdb-w7plus-avx2.7z Official UXP changes picked since my last build: - [DOM] Improve CC/WrapperCache for nursery objects. (a126257d80) - [DOM] Ensure the GC sees consistent states while adopting DOM nodes. (0ca2e3ea42) - [image] Hold a strong ref on error. (1a73ec0124) - [NSS] Improve input validation in DSAU signature decoding (d6ff046d72) - Bug 2025976 - Make txxPathNode be a value instead of reference. (f568fa4576) - [DOM/XSLT] Re-fetch index entry. (0bf878118a) - Bug 2025971 - Use txSingleNodeContext getter of owned txXPathNode. (368694e27a) - [webrtc] Fix off-by-one comparison. (8a4c6d11e0) - [gfx] Don't return dummy surfaces. (483bf0f43a) - [accessibility] Hold a strong ref to mParent in AccIterator. (b2dc9706ca) - Bug 2026293: Block scripts when initiating async scrollbar activity. (efbe3cb909) - [widget] Improve clearing of data during clipboard operations. (8c3c8b4e91) - [DOM] Add a SequenceRooter to ConsoleProfile. (470d5bd180) - [DOM] Add a death grip and use locals in BaseMediaResource::ModifyLoadFlags. (d8b8649f92) - [DOM/WebCrypto] Reject absurdly large salt values. (59887ccee8) - [DOM] Avoid leaks in nsXMLPrettyPrinter. (bee96124c5) - [NSS] Reject excessively large ASN.1 SEQUENCE OF in quickder. (9dd4cfcc64) - No issue - Deduplicate HTMLAllCollection::GetSupportedNames code. (c087ec3c00) - [DOM/XSLT] Prevent circular indexing (a0587881d6) - [NSS] Deep copy profile data in CERT_FindSMimeProfile. (5f41bdffd4) - Bug 2025483 update mBeginProcessing even on speex_resampler_set_rate() error (6cf972e912) - Bug 2025332 - ErrorResult should call ClearUnionData more. (be70077b66) - [gfx] Stop using binary arithmetic in software filtering. (4139582955) - [DOM/XSLT] Don't use txXPathNode by reference. (853057cf33) - Bug 2025370 - Properly handle error case in EventListenerManager::GetTypedEventHandler (096dffbcab) - Issue #3030 - Reduce layout stalls due to compositor hang-ups. (58f5dd61ae) - [js] Only allow strings as function name. (bcd936e548) - [NSS] Clarify extension negotiation mechanism for TLS Handshakes. (e4fd587adf) - Bug 2025973 - Use value-semantics for txExecutionState::TemplateRule (d419d35337) - [gfx] Port libyuv upstream bounds check. (99c241d56f) - [gfx] Explicitly cast calculation to int32. (602a2e35d8) - Bug 2029427 - Fix format string (3d9478ef97) - Bug 2029423 - Reset post_proc state. (a246b84e8b) - Bug 2029317 - Create RareArgumentsData before mutating the arguments object in MappedArgumentsObject::obj_defineProperty. (5858cd226c) - Bug 2029291. (5a065b86fc) - [security] Ensure `DispatchToMainThreadAndWait` isn't interrupted. (69bd6af877) - Bug 2029462 - store email on subject cache_entry in NSS trust domain. (7937766f4a) - No issue - Update our in-tree brotli library to 1.2.0+ (1020e0cd4f) - Revert "[gfx] Use local statics for harfbuzz callback pointers" (1f95ec61d6) - Merge remote-tracking branch 'origin/tracking' into custom (6be945e9da) - Bug 2030135 - improve error handling in PK11_ImportPrivateKeyInfoAndReturnKey (a8571576a1) - [DOM/media] Use extended channel data as-appropriate. (21d1a7dc24) - [DOM] Add some death grips to TextTrackManager. (f897efd802) - [image] Hold a strong ref on mInnerImage when requesting a refresh. (abd19068fb) - Bug 2029425 - Heap use-after-free in cert_VerifyCertChainOld via dangling certsList[] entry on NameConstraints violation. (aad6fdd59b) - Bug 2029323 - Improve size calculations in CMS content buffering (0329082cf0) - [js] Use isElement in UnmappedArgSetter to check if the argument was marked as deleted. (8a3427845f) - [gfx] OTS/HB: Correct bounds checks. (80b0fe5547) - Bug 2029727: Update RegExpStatics data after realloc (08f1b381fc) - [image] Don't do unnecessary work if the image is being destroyed. (85e32b8238) - [media] Spot-fix for ffvpx handling of bad video data. (a7a5092b23) - [parser] Handle foster parenting properly. (69d0209956) - Revert "[DOM] Ensure the GC sees consistent states while adopting DOM nodes." (dcf89d342f) - Bug 2027433 Treat null input channels as zero (83ef39ff44) - Bug 2027433 use nullptr for silent up-mix channels (42f1ba1d29) - Bug 2029806: Handle ill-formed UTF16 (f45cce8396) - [DOM/media] Add some sanity checks to MP3 demuxer. (c748294a4a) - [DOM] Validate size in FileReader. (dceb1c4fe6) - Issue #3066 - Follow-up: Reserve threads for burst use cases. (38dc7b394e) - [DOM] Do not interrupt worker execution if already scheduled. (bf526d7076) - [gfx] Add sanity checks in ANGLE D3D11 renderer. (208e06dc84) No official Pale-Moon changes picked since my last build. No official Basilisk changes picked since my last build. My changes picked since my last build: - js-HelperThreads: reduce maxParseThreads() to cpuCount - 1, or 1 if CPU is dual core or less. (04e063ddee) - dom/media: ffmpeg-audio: apply same changes to !MOZ_SAMPLE_TYPE_FLOAT32 code paths (27c36d1ea9) - Revert "Spoof to Firefox 128 for WebExtensions" (12e7d147cf) Update Notice: - You may delete file named icudt*.dat and icu63.dll inside program folder when updating from old releases. * Notice: From now on, UXP rev will point to `custom` branch of my UXP repo instead of MCP UXP repo, while "official UXP changes" shows only `tracking` branch changes. 5
roytam1 Posted May 1 Author Posted May 1 New build of BOC/UXP for XP! Test binary: MailNews Win32 https://o.rthost.win/boc-uxp/mailnews.win32-20260502-40a79c75-uxp-9161cd3bdb-xpmod.7z BNavigator Win32 https://o.rthost.win/boc-uxp/bnavigator.win32-20260502-40a79c75-uxp-9161cd3bdb-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/boc-uxp/tree/custom * Notice: the profile prefix (i.e. parent folder names) are also changed since 2020-08-15 build, you may rename their names before using new binaries when updating from builds before 2020-08-15. -- New build of HBL-UXP for XP! Test binary: IceDove-UXP(mail) https://o.rthost.win/hbl-uxp/icedove.win32-20260502-id-656ea98-uxp-9161cd3bdb-xpmod.7z IceApe-UXP(suite) https://o.rthost.win/hbl-uxp/iceape.win32-20260502-id-656ea98-ia-c642e3c-uxp-9161cd3bdb-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/icedove-uxp/tree/winbuild https://github.com/roytam1/iceape-uxp/tree/winbuild 2
roytam1 Posted May 1 Author Posted May 1 New build of post-deprecated Serpent/moebius for XP! * Notice: This repo will not be built on regular schedule, and changes are experimental as usual. ** Current moebius patch level should be on par with 52.9, but some security patches can not be applied/ported due to source milestone differences between versions. Test binary: Win32 https://o.rthost.win/basilisk/basilisk55-win32-git-20260502-6359349a8-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk55-win64-git-20260502-6359349a8-xpmod.7z repo: https://github.com/roytam1/basilisk55 Repo changes: - js-HelperThreads: reduce maxParseThreads() to cpuCount - 1, or 1 if CPU is dual core or less. (f57dec529) - import from UXP: [DOM] Improve CC/WrapperCache for nursery objects. (a126257d) (b7384e984) - import from UXP: [DOM] Ensure the GC sees consistent states while adopting DOM nodes. (0ca2e3ea) (028fc21c4) - import from UXP: [image] Hold a strong ref on error. (1a73ec01) (cd6e0c85c) - import from UXP: [NSS] Improve input validation in DSAU signature decoding (d6ff046d) (94fb8c8a6) - import from UXP: Bug 2025976 - Make txxPathNode be a value instead of reference. (f568fa45) (6dff65471) - import from UXP: [DOM/XSLT] Re-fetch index entry. (0bf87811) (f5b7cc8f4) - import from UXP: Bug 2025971 - Use txSingleNodeContext getter of owned txXPathNode. (368694e2) (48085b1b1) - import from UXP: [webrtc] Fix off-by-one comparison. (8a4c6d11) (45313bc26) - import from UXP: [gfx] Don't return dummy surfaces. (483bf0f4) (f6e88778a) - import from UXP: [accessibility] Hold a strong ref to mParent in AccIterator. (b2dc9706) (54630631b) - import from UXP: Bug 2026293: Block scripts when initiating async scrollbar activity. (efbe3cb9) (552fd9cd5) - import from UXP: [widget] Improve clearing of data during clipboard operations. (8c3c8b4e) (3d5310370) - import from UXP: [DOM] Add a SequenceRooter to ConsoleProfile. (470d5bd1) (567dd94b9) - import from UXP: [DOM] Add a death grip and use locals in BaseMediaResource::ModifyLoadFlags. (d8b8649f) (5d45ef230) - ported from UXP: [DOM/WebCrypto] Reject absurdly large salt values. (59887cce) (5eb1e866c) - import from UXP: [DOM] Avoid leaks in nsXMLPrettyPrinter. (bee96124) (5f5e71842) - import from UXP: [NSS] Reject excessively large ASN.1 SEQUENCE OF in quickder. (9dd4cfcc) (8679b848c) - import from UXP: No issue - Deduplicate HTMLAllCollection::GetSupportedNames code. (c087ec3c) (c85450456) - import from UXP: [DOM/XSLT] Prevent circular indexing (a0587881) (0d91cf9e1) - import from UXP: [NSS] Deep copy profile data in CERT_FindSMimeProfile. (5f41bdff) (028ac906b) - import from UXP: Bug 2025483 update mBeginProcessing even on speex_resampler_set_rate() error (6cf972e9) (7c1504d7a) - import from UXP: Bug 2025332 - ErrorResult should call ClearUnionData more. (be70077b) (893fa710a) - import from UXP: [gfx] Stop using binary arithmetic in software filtering. (41395829) (1baea0354) - import from UXP: [DOM/XSLT] Don't use txXPathNode by reference. (853057cf) (04f036803) - import from UXP: Bug 2025370 - Properly handle error case in EventListenerManager::GetTypedEventHandler (096dffbc) (4980c3894) - import from UXP: Issue #3030 - Reduce layout stalls due to compositor hang-ups. (58f5dd61) (7fbc7b10f) - import from UXP: [js] Only allow strings as function name. (bcd936e5) (42ef7b821) - import from UXP: [NSS] Clarify extension negotiation mechanism for TLS Handshakes. (e4fd587a) (4378c0b3c) - import from UXP: Bug 2025973 - Use value-semantics for txExecutionState::TemplateRule (d419d353) (87d44e0bd) - import from UXP: [gfx] Port libyuv upstream bounds check. (99c241d5) (aff4c6aaf) - import from UXP: [gfx] Explicitly cast calculation to int32. (602a2e35) (0b3db11cd) - import from UXP: Bug 2029427 - Fix format string (3d9478ef) (f66af5be4) - import from UXP: Bug 2029423 - Reset post_proc state. (a246b84e) (09653f2d4) - import from UXP: Bug 2029317 - Create RareArgumentsData before mutating the arguments object in MappedArgumentsObject::obj_defineProperty. (5858cd22) (fd5eeca9c) - import from UXP: Bug 2029291. (5a065b86) (354483e5c) - import from UXP: [security] Ensure `DispatchToMainThreadAndWait` isn't interrupted. (69bd6af8) (a057cae18) - import from UXP: Bug 2029462 - store email on subject cache_entry in NSS trust domain. (7937766f) (df3e0635e) - ported from UXP: No issue - Update our in-tree brotli library to 1.2.0+ (1020e0cd) (d22e189a4) - import from UXP: Revert "[gfx] Use local statics for harfbuzz callback pointers" (1f95ec61) (4b46a260f) - import from UXP: Bug 2030135 - improve error handling in PK11_ImportPrivateKeyInfoAndReturnKey (a8571576) (4dc05471f) - ported from UXP: [DOM/media] Use extended channel data as-appropriate. (21d1a7dc) (43f62a272) - ported from UXP: [DOM] Add some death grips to TextTrackManager. (f897efd8) (8fd3ea341) - import from UXP: [image] Hold a strong ref on mInnerImage when requesting a refresh. (abd19068) (52c7f4c2c) - import from UXP: Bug 2029425 - Heap use-after-free in cert_VerifyCertChainOld via dangling certsList[] entry on NameConstraints violation. (aad6fdd5) (a81abc48d) - import from UXP: Bug 2029323 - Improve size calculations in CMS content buffering (0329082c) (6ffc4f152) - import from UXP: [js] Use isElement in UnmappedArgSetter to check if the argument was marked as deleted. (8a342784) (72dd26b2f) - import from UXP: [gfx] OTS/HB: Correct bounds checks. (80b0fe55) (9de8f8123) - import from UXP: Bug 2029727: Update RegExpStatics data after realloc (08f1b381) (8d925fbd1) - import from UXP: [image] Don't do unnecessary work if the image is being destroyed. (85e32b82) (b2acfc720) - import from UXP: [media] Spot-fix for ffvpx handling of bad video data. (a7a5092b) (9ba9b9a29) - import from UXP: [parser] Handle foster parenting properly. (69d02099) (284e76377) - import from UXP: Revert "[DOM] Ensure the GC sees consistent states while adopting DOM nodes." (dcf89d34) (3149e2dd8) - import from UXP: Bug 2027433 Treat null input channels as zero (83ef39ff) (403ec902a) - import from UXP: Bug 2027433 use nullptr for silent up-mix channels (42f1ba1d) (ddf51d3fa) - Revert "ported from `custom` branch of UXP: Spoof to Firefox 128 for WebExtensions (ae7c40d4)" (2ec5548c9) - import from UXP: Bug 2029806: Handle ill-formed UTF16 (f45cce83) (feeff479f) - import from UXP: [DOM/media] Add some sanity checks to MP3 demuxer. (c748294a) (43464120e) - import from UXP: [DOM] Validate size in FileReader. (dceb1c4f) (96709570c) - import from UXP: [DOM] Do not interrupt worker execution if already scheduled. (bf526d70) (b79b897b6) - ported from UXP: [gfx] Add sanity checks in ANGLE D3D11 renderer. (208e06dc) (6359349a8) 3
raddy Posted May 2 Posted May 2 11 hours ago, roytam1 said: NM28XP build: Win32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260502-d849524bd-uxp-9161cd3bdb-xpmod.7z Next "regression" - https://midnightscene.cc/login if i press F12, i start to see a part of login form ps. palemoon-28.10.7a1.win32-git-20260418-d849524bd-uxp-d4c4c1f6ec-xpmod.7z & current basilisks have the same problem. 1
roytam1 Posted May 2 Author Posted May 2 2 hours ago, raddy said: Next "regression" - https://midnightscene.cc/login if i press F12, i start to see a part of login form ps. palemoon-28.10.7a1.win32-git-20260418-d849524bd-uxp-d4c4c1f6ec-xpmod.7z & current basilisks have the same problem. does any earlier version works? to me, its "starlight" effect canvas doesn't have proper z-index. setting it with `z-index:-1` should be shown as desired?
raddy Posted May 2 Posted May 2 2 hours ago, roytam1 said: does any earlier version works? without "starlight" palemoon-27.10.0.win32-git-20230107-61a988a8ac-xpmod.7z palemoon-27.10.0.win32-git-20220101-cbefc77b3-xpmod.7z palemoon-27.9.7.win32-git-20200104-fa517dfc6-xpmod.7z & FF 52 )
roytam1 Posted May 2 Author Posted May 2 18 minutes ago, raddy said: without "starlight" palemoon-27.10.0.win32-git-20230107-61a988a8ac-xpmod.7z palemoon-27.10.0.win32-git-20220101-cbefc77b3-xpmod.7z palemoon-27.9.7.win32-git-20200104-fa517dfc6-xpmod.7z & FF 52 ) lol thats too far away. and UXP build that works?
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now