1. 26 3月, 2015 1 次提交
  2. 24 3月, 2015 1 次提交
  3. 06 3月, 2015 1 次提交
  4. 13 2月, 2015 1 次提交
  5. 10 2月, 2015 1 次提交
  6. 20 1月, 2015 1 次提交
  7. 15 1月, 2015 1 次提交
    • T
      ALSA: hda - Select INPUT for Realtek HD-audio codec · ffcd28d8
      Takashi Iwai 提交于
      The commit commit [33f4acd3: ALSA: hda - Enable mic mute hotkey
      and LEDs for an HP machine] introduced a quirk for a HP machine
      involving with the input event handling.  Although the relevant code
      is protected via IS_ENABLED(CONFIG_INPUT), this doesn't suffice when
      the audio driver is built in while the input is module.
      
      As an easy workaround, this patch forcibly selects CONFIG_INPUT in
      Kconfig.  This shouldn't be a practical problem since CONFIG_INPUT is
      almost mandatory for all systems.  Also, this allows to remove the
      ugly ifdefs in the code.
      
      Fixes: 33f4acd3 ('ALSA: hda - Enable mic mute hotkey and LEDs for an HP machine')
      Acked-by: NDavid Henningsson <david.henningsson@canonical.com>
      Signed-off-by: NTakashi Iwai <tiwai@suse.de>
      ffcd28d8
  8. 11 1月, 2015 1 次提交
  9. 07 1月, 2015 1 次提交
  10. 19 12月, 2014 1 次提交
  11. 17 12月, 2014 2 次提交
  12. 06 12月, 2014 1 次提交
  13. 02 12月, 2014 1 次提交
  14. 21 11月, 2014 2 次提交
  15. 20 11月, 2014 1 次提交
  16. 18 11月, 2014 2 次提交
  17. 12 11月, 2014 1 次提交
    • K
      ALSA: hda/realtek - Change EAPD to verb control · 394c97f8
      Kailang Yang 提交于
      This will fix no sound in Linux system after reboot from windows.
      
      Change log:
      - alc662_fill_coef() is replaced with alc_fill_eapd_coef_idx()
        and move into alc_auto_init_amp().
      - For ALC262, ALC267, ALC268, ALC269, ALC233, ALC255, ALC280, ALC282,
        ALC283, ALC284, ALC285, ALC286, ALC288, ALC290, ALC292, ALC293, ALC294,
        ALC668, ALC888VC, ALC888VD, ALC891, ALC892, ALC898 and ALC1150, add update
        COEF control for EAPD setting.
      - Remove alc269_fill_coef() for update EAPD control line.
      
      ADDITIONAL NOTE:
      Many Realtek cdoecs have a COEF bit to switch the master amp control
      between COEF and EAPD.  Windows drivers seem using COEF while we use
      EAPD, which is more standard.  As a result, some system suffer from
      the silent output when booting after Windows.  This patch sets the
      COEF bits on the relevant codecs properly to switch to EAPD control.
      
      Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=87771Signed-off-by: NKailang Yang <kailang@realtek.com>
      Signed-off-by: NTakashi Iwai <tiwai@suse.de>
      394c97f8
  18. 06 11月, 2014 1 次提交
  19. 05 11月, 2014 1 次提交
  20. 30 10月, 2014 4 次提交
  21. 27 10月, 2014 1 次提交
  22. 24 10月, 2014 2 次提交
  23. 20 10月, 2014 2 次提交
  24. 08 10月, 2014 1 次提交
  25. 07 10月, 2014 3 次提交
  26. 16 9月, 2014 1 次提交
    • T
      ALSA: hda - Allow multiple callbacks for jack · 1a4f69d5
      Takashi Iwai 提交于
      So far, hda_jack infrastructure allows only one callback per jack, and
      this makes things slightly complicated when a driver wants to assign
      multiple tasks to a jack, e.g. the standard auto-mute with a power
      up/down sequence.  This can be simplified if the hda_jack accepts
      multiple callbacks.
      
      This patch is such an extension: the callback-specific part (the
      function and private_data) is split to another struct from
      hda_jack_tbl, and multiple such objects can be assigned to a single
      hda_jack_tbl entry.
      
      The new struct hda_jack_callback is passed to each callback function
      now, thus the patch became bigger than expected.  But these changes
      are mostly trivial.
      Signed-off-by: NTakashi Iwai <tiwai@suse.de>
      1a4f69d5
  27. 11 9月, 2014 1 次提交
    • T
      ALSA: hda - Get rid of action field from struct hda_jack_tbl · 62f949bf
      Takashi Iwai 提交于
      The action value assigned to each hda_jack_tbl entry is mostly
      superfluous.  The actually used values are either the widget NID or a
      value specific to the callback.
      
      The former case can be simply replaced by a reference to widget NID
      itself.  The only place doing the latter is STAC/IDT codec driver for
      the powermap handling.  But, the code doesn't need to check the action
      field at all -- the function jack_update_power() is called either with
      a specific pin or with NULL.  So the check of jack->action can be
      removed completely there, too.
      Signed-off-by: NTakashi Iwai <tiwai@suse.de>
      62f949bf
  28. 03 9月, 2014 3 次提交