…LING_DOWN_FACTOR to 2, no need for such aggressive rate reduction now we have boostpulse. As CPU frequency is shoved to max everytime we interact, it negates the need to have aggressive ramping down of samples, doubling is enough. DEFAULT_FREQ_BOOST_TIME to 500ms, 1 second was unnecessarily high. 500ms is enough time for Android animations to complete.
… for A9s. They *do* idle efficiently at higher frequencies, especially with OMAP OSWR or T3 LP2 enabled.
For applications that require a one-shot increase in frequency, the boostpulse interface would help to scale to max for the time specified in boostime. This helps applications that currently increases scaling_min_freq and resets later after a fixed duration. Instead of changing scaling_min_freq applications can directly write to boostpulse and governer resets during regular timer after the time expires. Based on work by Narayanan Gopalakrishnan <email@example.com> with the following changes: Remove boostfreq tunable and set it to policy->max always Apply a sampling down factor of 10 and make the default boost time 1 sec. attributes sampling_rate & boostpulse to 666 mode to avoid ramdisk change.
…k_timer is pending before attempting to add it. If it is pending, use mod_timer to change the timeout. Adding regardless results in a BUG() and a subsequent panic.
Signed-off-by: Ezekeel <firstname.lastname@example.org>
… is invoked when the device is in early suspend. PM_MAX causes wifi speeds to drop dramatically, so not ideal for cifs, adb wireless or file transfers whilst the screen is off. If this works like the bcm4329 one, it won't cause additional battery usage.
…ts of Oopses occasionally with a NULL pointer dereference always in the same function. I believe in some circumstances *dev_id is NULL which is causing said issue, so check before use and bail if NULL.
commit 57494c162df513dcc04065acfb34b68cd8b8da68 Author: SIMOND François <email@example.com> Date: Mon Dec 19 12:55:33 2011 +0100 Voodoo Sound: update source code for ICS kernel sources commit ba7e0d3d2d42db104a10940ed1b373de9f5e5434 Author: SIMOND François <firstname.lastname@example.org> Date: Fri Jul 8 16:51:42 2011 +0200 Voodoo sound: driver v10 Improvement: - HW EQ support: smooth activation/deactivation and gain changes Bugfixes: - HP volume smoothing loop on low levels with negative digital offsets - wm8994_write logging on Nexus S New supported devices: - Galaxy Tab 7" Gingerbread Kernels support (based on M180S) - Galaxy Tab 10.1 - beta commit ced38fcf6fad9feb063272187542f605601df6ff Author: SIMOND François <email@example.com> Date: Wed Jun 15 16:30:14 2011 +0200 Voodoo sound: fix wm8994_write logging routine missing register/value for Nexus S-based kernels commit 5af820a283a0ad61886582051102cc7843659561 Author: SIMOND François <firstname.lastname@example.org> Date: Fri May 20 08:43:01 2011 +0200 Voodoo sound: driver v9 New features - advanced logging controllable via debug_log infos: 1, verbose: 2 - implements hardware EQ - implements hardware stereo expansion effect - new concept: digital_gain: makes room for effects with negative gains avoid saturation using hardware DRC as limiter act as compresser+limiter with positive gains (gain unit is mili-decibels, min -71625, max 36000) negative digital_gain are analog compensated if possible - super smooth headphone amp volume changes New supported device - Compatible with M110S Gingerbread kernel sources commit 72fdd557b06ac983cdb8412512bc696df2b7da75 Author: SIMOND François <email@example.com> Date: Mon Apr 25 06:49:14 2011 +0200 Voodoo sound: driver v8 - Fixes DAC direct not being applied in several situations - Fixes 1 channel only on Captivate headphone+mic calls - Add full support for Samsung Gingerbread kernels including complete FM radio tunings commit cd929a8f9046b4f1be8af2c9d090a0ad1b3dde9e Author: SIMOND François <firstname.lastname@example.org> Date: Mon Mar 28 07:05:04 2011 +0200 Voodoo sound: update to driver v7 - definite fix for DAC direct call issue - reshaped source code (Linux Coding Style!) - less hooks in pre-existing driver commit 9847bed9d3a9484bfdd72ee8ad5737bd317f8b0a Author: SIMOND François <email@example.com> Date: Thu Mar 24 09:10:43 2011 +0100 Voodoo sound: modularization-related change and path management improvements commit 681c3e013cc811a1cde65c1d54858c9ba59ffa0f Author: SIMOND François <firstname.lastname@example.org> Date: Mon Mar 21 06:25:43 2011 +0100 Voodoo sound: bump version to v6 commit cf36abb84b15dff206a1a4d19f0b7ef7c09bfa7f Author: SIMOND François <email@example.com> Date: Fri Mar 18 02:39:58 2011 +0100 Voodoo sound: apply speaker tuning control only when active path is speaker, fix reversed is_path(SPEAKER) logic commit 3a1659a99ca9f3cfa8ab5271b886940fdb7bd725 Author: SIMOND François <firstname.lastname@example.org> Date: Thu Mar 17 07:40:08 2011 +0100 Voodoo sound: update to driver v5 - New speaker tuning mode for music using different hardware EQ settings than stock ones - Microphone presets now supported on Nexus S. - Microphone presets balanced and loud reworked with lower background noise and higher dynamic range for balanced: -2dB noise, +6dB dynamic range. - New DAC direct option: reduces distortion and improves SNR by a small margin (bypass usless analog channel mixer) - Anti-jitter mode on by default. - Same exact source is now used on each device supported (still: only useful code is compiled in via ifdefs) - ability to fully disable/re-enable the driver and its sysfs interface: 0 or 1 in /sys/class/misc/voodoo_sound_control/enable - refactored source code commit d1d535e2fd3aaf359d181fd48762fe2e4a2c1cfb Author: SIMOND François <email@example.com> Date: Mon Jan 31 05:43:30 2011 +0100 Voodoo sound: driver v4 for Nexus S. Microphone presets feature deactivated: needs some fine adjustement for the code & hardware
Setting proper mcs rate mask allows to connect in 802.11n-only mode Change-Id: Iffd47a60c92a74d29e42d81d79fa64463f5895c5 Signed-off-by: Dmitry Shmidt <firstname.lastname@example.org>
In the past, a process could only see its own stats (uid-based summary, and details). Now we allow any process to see other UIDs uid-based stats, but still hide the detailed stats. Change-Id: I7666961ed244ac1d9359c339b048799e5db9facc Signed-off-by: JP Abgrall <email@example.com>
…ation. Signed-off-by: Ashish Sharma <firstname.lastname@example.org>
Change-Id: I1b785c47bda03208d396473a9d1c94631f5624d9 Signed-off-by: Mike Lockwood <email@example.com>
When USB was disconnected, we were freeing our audio_dev struct in audio_unbind before the audio system had cleaned up, resulting in a hang in audio_pcm_close. We now statically allocate the audio_dev struct to avoid this problem. Change-Id: I58ad21eaa20dcf4aa74ee614ef3b6ed2c91d52a1 Signed-off-by: Mike Lockwood <firstname.lastname@example.org>
Fixes watchdog reset on USB disconnect Larger packet size fixes performance in fullspeed mode Signed-off-by: Mike Lockwood <email@example.com>
isochronous endpoints are not limited to 64 bytes Signed-off-by: Mike Lockwood <firstname.lastname@example.org>
This is a merge of backported changes from kernel/exynos android-exynos-3.4 -- commit 62cb87fd5b32f8cf07bc46a691318f4aa2cfc07d i2c-s3c2410: clean up i2c irq after timeout Otherwise, later irq handler will try to access freed memory region and may crash. Change-Id: I0ab9157b6f0d48d3a413e59a27641e88cd9db2eb Signed-off-by: Jinhee Hyeon <email@example.com> Conflicts: drivers/i2c/busses/i2c-s3c2410.c -- commit cd668911107bd1432047406e31f3be7278dc4046 i2c-s3c2410: add missing stop sequence for I2C transfer To clearly terminate the connection, clear pending bit and serial out disable settings are added Change-Id: Ic5ccd80941a724f63ed7f22e777408725c5414ee Signed-off-by: HoJune Byun <firstname.lastname@example.org> -- commit 6aeac78519c91d50757dc492ff12fbabc4cc756b i2c-s3c2410: I2C connection termination forcely To forcely terminate the connection, Serial Out bit is disabled. In case of timeout, or HDMIPHY bug Change-Id: I29c9ef2b8c35c7c970cbef4d5a856e178794de6e Signed-off-by: HoJune Byun <email@example.com> -- commit 2951f354549bf65cf9a4733565841553fbc8a964 i2c-s3c2410: added STATE_STOP checking codes To prevent duplicated stop execution, added STATE_STOP checking codes before s3c24xx_i2c_stop function. Change-Id: I29417d4dbf94280e1d629338e85e54ad79e6be89 Signed-off-by: Taekgyun Ko <firstname.lastname@example.org> Conflicts: drivers/i2c/busses/i2c-s3c2410.c Change-Id: I78ea9b014ed9656d6b5c4628d36bc8edd668b8e0 Signed-off-by: JP Abgrall <email@example.com>