Changes between Version 22 and Version 23 of QtBackLog


Ignore:
Timestamp:
Oct 25, 2009 9:07:13 PM (11 years ago)
Author:
zecke@selfish.org
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • QtBackLog

    v22 v23  
    22
    33=== Qt problems ===
    4 ||Description||Branch||Landed in||
    5 ||ARM optimized qdrawhelper is only enabled for RCVT (symbian) . The  needs to be transferred to gcc inline assembly. This will optimize memfill, Source, SourceOver and some blend functions.|| || ||
    6 ||ARMv6 should be auto detected. Currently arm/armv6 can be passed as -embedded options. The benefit of ARMv6 in src/corelib is the use of atomic load and exchange, instead of the old "swap" extension. `__ARCH_ARM_6__` should be used to detect it.|| || ||
    7 ||Enabled pld (preload) in qdrawhelper for ARMv5te and upwards|| || ||
    8 ||Optimize scrolling of Windowsurfaces. Either optimize memcpy/memmove or consider using a tile based window surface|| || ||
    9 ||QImageReader deserves some optimatations. E.g. decode to the QImage provided to the decoder, do not parse every gif when trying to determine the size.||[http://qt.gitorious.org/~zecke/qt/zeckes-qt/commits/merge/qimagehandler-changes image-changes]|| ||
    10 ||Cut down on usage of QImage::scanLine inside Qt. This was already done in the GIF and PNG decoder.||[http://qt.gitorious.org/~zecke/qt/zeckes-qt/commits/merge/qimagehandler-changes image-changes]|| ||
    11 ||Make QGifHandler::imageCount scan through the images. This can give a 5% speedup in the image_cycling reduction.|| || ||
    12 ||Make picking a QImageIOHandler in QImageReader faster. Currently even the TIFF plugin is asked to handle images.|| || ||
    13 ||Consider creating a tile based QImage to allocate image chunks from an image pool with fixed size images|| || ||
    14 ||Bring zero-copy to QIODevice... big one. That is good for networking and image decoding.|| || ||
    15 ||Consider creating a SharedBuffer that works on tiles/segments|| || ||
    16 ||Changes to QHttpNetworkReply to reduce memory reallocations. Reserve some bytes instead of doing ''QByteArray::append'' all them time.|| || ||
    17 ||Changes to QNetworkReply to remove quadratic runtime in QNetworkReplyHandler and latency fixes||[http://qt.gitorious.org/~zecke/qt/zeckes-qt/commits/merge/network-changes network-changes] || ||
    18 ||Cut down on the usage of QUrl::toEncoded as it shows up in the profile of starting jobs.|| || ||
    19 ||QBENCHMARK results are hard to analyze and compare. We need a single and simple way to say if something is faster or slower. Be a bit like sunpsider, do the same math as well, mention slowest and fastest run.|| || ||
     4||Description||Branch||Landed in||Landed Date||
     5||ARM optimized qdrawhelper is only enabled for RCVT (symbian) . The  needs to be transferred to gcc inline assembly. This will optimize memfill, Source, SourceOver and some blend functions.|| || || ||
     6||ARMv6 should be auto detected. Currently arm/armv6 can be passed as -embedded options. The benefit of ARMv6 in src/corelib is the use of atomic load and exchange, instead of the old "swap" extension. `__ARCH_ARM_6__` should be used to detect it.|| || || ||
     7||Enabled pld (preload) in qdrawhelper for ARMv5te and upwards|| || || ||
     8||Optimize scrolling of Windowsurfaces. Either optimize memcpy/memmove or consider using a tile based window surface|| || || ||
     9||QImageReader deserves some optimatations. E.g. decode to the QImage provided to the decoder, do not parse every gif when trying to determine the size.||[http://qt.gitorious.org/~zecke/qt/zeckes-qt/commits/merge/qimagehandler-changes image-changes]|| || ||
     10||Cut down on usage of QImage::scanLine inside Qt. This was already done in the GIF and PNG decoder.||[http://qt.gitorious.org/~zecke/qt/zeckes-qt/commits/merge/qimagehandler-changes image-changes]|| || ||
     11||Make QGifHandler::imageCount scan through the images. This can give a 5% speedup in the image_cycling reduction.|| || || ||
     12||Make picking a QImageIOHandler in QImageReader faster. Currently even the TIFF plugin is asked to handle images.|| || || ||
     13||Consider creating a tile based QImage to allocate image chunks from an image pool with fixed size images|| || || ||
     14||Bring zero-copy to QIODevice... big one. That is good for networking and image decoding.|| || || ||
     15||Consider creating a SharedBuffer that works on tiles/segments|| || || ||
     16||Changes to QHttpNetworkReply to reduce memory reallocations. Reserve some bytes instead of doing ''QByteArray::append'' all them time.|| || || ||
     17||Changes to QNetworkReply to remove quadratic runtime in QNetworkReplyHandler and latency fixes||[http://qt.gitorious.org/~zecke/qt/zeckes-qt/commits/merge/network-changes network-changes] || || ||
     18||Cut down on the usage of QUrl::toEncoded as it shows up in the profile of starting jobs.|| || || ||
     19||QBENCHMARK results are hard to analyze and compare. We need a single and simple way to say if something is faster or slower. Be a bit like sunpsider, do the same math as well, mention slowest and fastest run.|| || || ||
    2020
    2121=== QtWebKit problems ===
    22 ||Bug||Description||Commit||
    23 ||||Loading of http://en.wikipedia.org/wiki/Maxwell_equations is triggering bad "scheduling". The loading_standalone test has been extracted but is not able to highlight the problem. This is a Qt and QtWebKit issue, it must be understood first.||||
    24 ||||TCmalloc needs to be tested. Does it make anything faster/slower? How to test memory fragmentation.||||
    25 ||||The fuzz testing will make QtWebKit go in infinite loops and stop working. This situation needs to be analyzed as this can be a serve reliabality problem.||||
    26 ||[https://bugs.webkit.org/show_bug.cgi?id=30211 #30211]||Research a better QImage -> QPixmap migration strategy. Currently we convert every QImage to a QPixmap and this is bad for times where we need to draw a scaled version of the image or will only draw this image once.||||
    27 ||[https://bugs.webkit.org/show_bug.cgi?id=30293 #30293]||Loading speed regressed due image changes?||||
    28 ||[https://bugs.webkit.org/show_bug.cgi?id=30301 #30301]||Loading bug/API bug on loading amazon in QtEmbedded Linux. Loading does not complete.||||
    29 ||[https://bugs.webkit.org/show_bug.cgi?id=29279 #29279]||Decode images to RGB16 or smaller (instead of RGB32) to conserve memory. The patch introduces new API to Qt.||||
    30 ||||Go through the Palm changes and consider adopting them. Currently candidates are: Use stack/class memory in CSSParser, use JDCT_Fast for Jpeg decoding (to be achieved with QImageReader to set quality to 49), changes to Cached* to throw away encoded data more early. Palm avoids floating point operation on premultiplied alpha in setRGBA.||||
    31 ||[https://bugs.webkit.org/show_bug.cgi?id=30740 #30740]||Using QImageReader::setQuality(49) will use JDCT_IFAST and promises to give a 5% speedup on image_cycling.||||
    32 ||[https://bugs.webkit.org/show_bug.cgi?id=30210 #30210]||Create Embedded profile that will tune for embedded usage. This can mean to save memory in favor of speed, add a speed gain by avoiding certain operations, trading image quality for speed||||
    33 ||||Change CursorQt.cpp to not create QPixmap's right away but only on first load. This will be a minor memory improvement.||||
    34 ||[https://bugs.webkit.org/show_bug.cgi?id=27538 #27538]||~~Image decoder changes needs to be measured. Do they make anything faster/slower? Do they consume more or less memory?~~||[https://trac.webkit.org/changeset/49179 49179] [https://trac.webkit.org/changeset/49180 49180] [https://trac.webkit.org/changeset/49181 49181] [https://trac.webkit.org/changeset/49182 49182] [https://trac.webkit.org/changeset/49183 49183] [https://trac.webkit.org/changeset/49185 49185] [https://trac.webkit.org/changeset/49186 49186] [https://trac.webkit.org/changeset/49559 49559]||
     22||Bug||Description||Commit||Date||
     23||||Loading of http://en.wikipedia.org/wiki/Maxwell_equations is triggering bad "scheduling". The loading_standalone test has been extracted but is not able to highlight the problem. This is a Qt and QtWebKit issue, it must be understood first.||||||
     24||||TCmalloc needs to be tested. Does it make anything faster/slower? How to test memory fragmentation.||||||
     25||||The fuzz testing will make QtWebKit go in infinite loops and stop working. This situation needs to be analyzed as this can be a serve reliabality problem.||||||
     26||[https://bugs.webkit.org/show_bug.cgi?id=30211 #30211]||Research a better QImage -> QPixmap migration strategy. Currently we convert every QImage to a QPixmap and this is bad for times where we need to draw a scaled version of the image or will only draw this image once.||||||
     27||[https://bugs.webkit.org/show_bug.cgi?id=30293 #30293]||Loading speed regressed due image changes?||||||
     28||[https://bugs.webkit.org/show_bug.cgi?id=30301 #30301]||Loading bug/API bug on loading amazon in QtEmbedded Linux. Loading does not complete.||||||
     29||[https://bugs.webkit.org/show_bug.cgi?id=29279 #29279]||Decode images to RGB16 or smaller (instead of RGB32) to conserve memory. The patch introduces new API to Qt.||||||
     30||||Go through the Palm changes and consider adopting them. Currently candidates are: Use stack/class memory in CSSParser, use JDCT_Fast for Jpeg decoding (to be achieved with QImageReader to set quality to 49), changes to Cached* to throw away encoded data more early. Palm avoids floating point operation on premultiplied alpha in setRGBA.||||||
     31||[https://bugs.webkit.org/show_bug.cgi?id=30740 #30740]||Using QImageReader::setQuality(49) will use JDCT_IFAST and promises to give a 5% speedup on image_cycling.||||||
     32||[https://bugs.webkit.org/show_bug.cgi?id=30210 #30210]||Create Embedded profile that will tune for embedded usage. This can mean to save memory in favor of speed, add a speed gain by avoiding certain operations, trading image quality for speed||||||
     33||||Change CursorQt.cpp to not create QPixmap's right away but only on first load. This will be a minor memory improvement.||||||
     34||[https://bugs.webkit.org/show_bug.cgi?id=27538 #27538]||~~Image decoder changes needs to be measured. Do they make anything faster/slower? Do they consume more or less memory?~~||[https://trac.webkit.org/changeset/49179 49179] [https://trac.webkit.org/changeset/49180 49180] [https://trac.webkit.org/changeset/49181 49181] [https://trac.webkit.org/changeset/49182 49182] [https://trac.webkit.org/changeset/49183 49183] [https://trac.webkit.org/changeset/491865 49185] [https://trac.webkit.org/changeset/49186 49186] [https://trac.webkit.org/changeset/49559 49559]||6.10.2009-14.10.2009||