Send commitlog mailing list submissions to
commitlog@lists.openmoko.org
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.openmoko.org/mailman/listinfo/commitlog
or, via email, send a message with subject or body 'help' to
[EMAIL PROTECTED]
You can reach the person managing the list at
[EMAIL PROTECTED]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of commitlog digest..."
Today's Topics:
1. Holger's qtopia repo: Changes to 'master' ([EMAIL PROTECTED])
2. Holger's qtopia repo: Changes to 'master' ([EMAIL PROTECTED])
3. Holger's qtopia repo: Changes to 'master' ([EMAIL PROTECTED])
4. Holger's qtopia repo: Changes to 'master' ([EMAIL PROTECTED])
5. development kernel tree: Changes to 'andy' ([EMAIL PROTECTED])
--- Begin Message ---
src/libraries/qtopia/themedview.cpp | 132 ++++++++++++++++++++++++++++++-----
src/libraries/qtopia/themedview.h | 1 +
2 files changed, 116 insertions(+), 17 deletions(-)
New commits:
commit 876ed02211e50547856491ad54a75299b62efe4c
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 13:50:35 2008 +0200
[theme] Correct the signature and do not force a layout on template items
It is a Qt 2/3'ish API and I passed the wrong number of arguments to it.
AllowLayout was actually not passed but state, fix that and have one
layout call less.
commit 9917d53a1505d8b297d085a664b23b4ee95e567c
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 13:49:13 2008 +0200
[theme] Collect the various updates that will come and then do a layout
Collect all updates and then do only one layout for the changes
to the states. E.g. with CallScreen we do change more than one item
and waiting for the new layout is making a difference.
commit 311ca4fd3bb1dee01db6aac5373d26acee0c02f2
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 12:33:04 2008 +0200
[theme] Huge performance boost on Qt/X11. Do not rescale images to 0,0
When a ThemeItem becomes invisible the width/height of the image gets
0,0
and we attempt to scale to that resolution. If we encounter a 0,0 scale
request just ignore it and assume that once the item is visible again
the
old size gets reused. This is a huge performance boost for the
CallScreen.
commit 4417b763eefaf586255e588c2a82a94bb58afd72
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 11:49:22 2008 +0200
[theme] Add ENTRY/EXIT to each function to assist in performance analysis
commit 0c2ea423d9ae93f1b302e774049a242a0a24b25e
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 10:00:35 2008 +0200
[theme] Do not allow to layout on mouse press/release
Do not allow to layout an item on mouse move/release for performance
reasons.
--- End Message ---
--- Begin Message ---
src/libraries/qtopia/themedview.cpp | 132 ++++++++++++++++++++++++++++++-----
src/libraries/qtopia/themedview.h | 1 +
2 files changed, 116 insertions(+), 17 deletions(-)
New commits:
commit 876ed02211e50547856491ad54a75299b62efe4c
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 13:50:35 2008 +0200
[theme] Correct the signature and do not force a layout on template items
It is a Qt 2/3'ish API and I passed the wrong number of arguments to it.
AllowLayout was actually not passed but state, fix that and have one
layout call less.
commit 9917d53a1505d8b297d085a664b23b4ee95e567c
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 13:49:13 2008 +0200
[theme] Collect the various updates that will come and then do a layout
Collect all updates and then do only one layout for the changes
to the states. E.g. with CallScreen we do change more than one item
and waiting for the new layout is making a difference.
commit 311ca4fd3bb1dee01db6aac5373d26acee0c02f2
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 12:33:04 2008 +0200
[theme] Huge performance boost on Qt/X11. Do not rescale images to 0,0
When a ThemeItem becomes invisible the width/height of the image gets
0,0
and we attempt to scale to that resolution. If we encounter a 0,0 scale
request just ignore it and assume that once the item is visible again
the
old size gets reused. This is a huge performance boost for the
CallScreen.
commit 4417b763eefaf586255e588c2a82a94bb58afd72
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 11:49:22 2008 +0200
[theme] Add ENTRY/EXIT to each function to assist in performance analysis
commit 0c2ea423d9ae93f1b302e774049a242a0a24b25e
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 10:00:35 2008 +0200
[theme] Do not allow to layout on mouse press/release
Do not allow to layout an item on mouse move/release for performance
reasons.
--- End Message ---
--- Begin Message ---
src/libraries/qtopia/themedview.cpp | 10 ++-
.../phone/callhistory/default/callhistory.cpp | 72 +++++++++-----------
src/server/phone/callhistory/default/callhistory.h | 5 +-
3 files changed, 43 insertions(+), 44 deletions(-)
New commits:
commit 70e8c35087f6d13ea84a4f502b3c9ef821be3623
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Tue Jul 1 08:20:13 2008 +0200
[theme] If we layout in the paintEvent then ignore the clipping
If we have to layout consider the whole screen dirty and paint
everything. Block updates as we are going to draw everything anyway.
This makes a visual difference on X11
commit 082e8f101b132c8f35e1e756540f782d7c865e85
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Tue Jul 1 08:19:48 2008 +0200
[theme] Disable printing performance numbers
commit 1774ed5b333d265971eea39f2ba1105c5ac22a09
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 19:49:27 2008 +0200
[callhistory] Elide the number label as well
It is quite tricky to layout the elided text properly. If the text is
too
big the QFormLayout will wrap our line. So taking the ->x() position of
our
label into account will lead to different results depending on what you
displayed
before (long text, short text). To avoid this take the x() + width() of
our
"label" into account and layout in this regard. With the QFormLayout it
is
quite easy to grow out of the screen which makes a class like
ElidedLabel
really hard to do. This is the best I can get it for now. With some
more time
a custom boxlayout class and elided label class would be the way to go.
(Bug #1504)
commit 389272760c11f3211b0a4025682bb0efad59dae0
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 19:03:12 2008 +0200
[callhistory] Update the widget directly instead of the event filter hooks
Update the content of the CallHistoryView directly instead of going
through the event filter. The root cause might be that the ::update
method of QWidget was never called from the reimplementation? who knows.
commit 2d902ac4dc2ccc48649ef17215d189d1d3e55d24
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 17:54:45 2008 +0200
[callhistory] Updating it once should be enough
Decided to use the activated signal only. Seems to work with
mouse and other means of input.
commit 5ccc621d92b27a9fba3bd62bdaa7139ce99bd512
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 17:30:44 2008 +0200
[x11] Small oversight and mark the window properly
This was not a real issue as we have our fallback that will
mark the window on show.
--- End Message ---
--- Begin Message ---
src/libraries/qtopia/themedview.cpp | 10 ++-
.../phone/callhistory/default/callhistory.cpp | 72 +++++++++-----------
src/server/phone/callhistory/default/callhistory.h | 5 +-
3 files changed, 43 insertions(+), 44 deletions(-)
New commits:
commit 70e8c35087f6d13ea84a4f502b3c9ef821be3623
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Tue Jul 1 08:20:13 2008 +0200
[theme] If we layout in the paintEvent then ignore the clipping
If we have to layout consider the whole screen dirty and paint
everything. Block updates as we are going to draw everything anyway.
This makes a visual difference on X11
commit 082e8f101b132c8f35e1e756540f782d7c865e85
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Tue Jul 1 08:19:48 2008 +0200
[theme] Disable printing performance numbers
commit 1774ed5b333d265971eea39f2ba1105c5ac22a09
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 19:49:27 2008 +0200
[callhistory] Elide the number label as well
It is quite tricky to layout the elided text properly. If the text is
too
big the QFormLayout will wrap our line. So taking the ->x() position of
our
label into account will lead to different results depending on what you
displayed
before (long text, short text). To avoid this take the x() + width() of
our
"label" into account and layout in this regard. With the QFormLayout it
is
quite easy to grow out of the screen which makes a class like
ElidedLabel
really hard to do. This is the best I can get it for now. With some
more time
a custom boxlayout class and elided label class would be the way to go.
(Bug #1504)
commit 389272760c11f3211b0a4025682bb0efad59dae0
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 19:03:12 2008 +0200
[callhistory] Update the widget directly instead of the event filter hooks
Update the content of the CallHistoryView directly instead of going
through the event filter. The root cause might be that the ::update
method of QWidget was never called from the reimplementation? who knows.
commit 2d902ac4dc2ccc48649ef17215d189d1d3e55d24
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 17:54:45 2008 +0200
[callhistory] Updating it once should be enough
Decided to use the activated signal only. Seems to work with
mouse and other means of input.
commit 5ccc621d92b27a9fba3bd62bdaa7139ce99bd512
Author: Holger Freyther <[EMAIL PROTECTED]>
Date: Mon Jun 30 17:30:44 2008 +0200
[x11] Small oversight and mark the window properly
This was not a real issue as we have our fallback that will
mark the window on show.
--- End Message ---
--- Begin Message ---
Rebased ref, commits from common ancestor:
commit 18446cfdf099cc551b91afcf5043bf16847bd08b
Author: Andy Green <[EMAIL PROTECTED]>
Date: Tue Jul 1 09:28:18 2008 +0100
test-touchscreen-pressure.patch
Signed-off-by: Andy Green <[EMAIL PROTECTED]>
commit e9ea4cdbc1dd475fafd2f49217d0c993686adce3
Author: Matt <[EMAIL PROTECTED]>
Date: Tue Jul 1 09:28:18 2008 +0100
add-ar6k-wake-interrupt.patch
Signed-off-by: Matt Hsu <[EMAIL PROTECTED]>
- add an interrupt for ar6k wifi module
commit 931647795f9cb799717ea9ded5ebc31432b89e97
Author: Andy Green <[EMAIL PROTECTED]>
Date: Tue Jul 1 09:28:18 2008 +0100
test-touchscreen-median.patch
Signed-off-by: Andy Green <[EMAIL PROTECTED]>
commit c27edf64d78abb0bb8d3004a231b6a2fb4c9931e
Author: Andrzej Zaborowski <[EMAIL PROTECTED]>
Date: Tue Jul 1 09:28:17 2008 +0100
From 119f4e02ba81cffe4dbc88d8ff667048ad28d925 Mon Sep 17 00:00:00 2001
Subject: [PATCH] Hacky CONFIG_NO_IDLE_HZ (dyn-tick) support for S3C24xx.
commit d10f5068c88a2a190c347bf2a78cb8604a91af83
Author: Andy Green <[EMAIL PROTECTED]>
Date: Tue Jul 1 09:28:17 2008 +0100
introduce-resume-exception-capture.patch
This patch introduces a new resume debugging concept: if we
get an OOPS inbetween starting suspend and finishing resume, it
uses a new "emergency spew" device similar to BUT NOT REQUIRING
CONFIG_DEBUG_LL to dump the syslog buffer and then the OOPS
on the debug device defined by the existing CONFIG_DEBUG_S3C_UART
index. But neither CONFIG_DEBUG_LL nor the S3C low level configs
are needed to use this feature.
Another difference between this feature and CONFIG_DEBUG_LL is that
it does not affect resume timing, ordering or UART traffic UNLESS
there is an OOPS during resume.
The patch adds three global exports, one to say if we are inside
suspend / resume, and two callbacks for printk() to use to init
and dump the emergency data. The callbacks are set in s3c serial
device init, but the whole structure is arch independent.
Signed-off-by: Andy Green <[EMAIL PROTECTED]>
--- End Message ---
_______________________________________________
commitlog mailing list
commitlog@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/commitlog
