Merge tag 'asoc-fix-v5.19-rc0' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Fixes for v5.19

A few more fixes that came in during the merge window - nothing
huge here, there is one core fix for DPCM from Pierre but mostly
driver changes.

Follow the pattern used by other selftests like memfd and fall back on the
standard toolchain options to build with a system installed alsa-lib if
we don't get anything from pkg-config. This reduces our build dependencies
a bit in the common case while still allowing use of pkg-config in case
there is a need for it.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20220531151337.2933810-1-broonie@kernel.orgSigned-off-by: Takashi Iwai <tiwai@suse.de>

Maris found out that the quirk for TEAC devices to work around the
clock setup is needed to apply only when the base clock is changed,
e.g. from 48000-based clocks (48000, 96000, 192000, 384000) to
44100-based clocks (44100, 88200, 176400, 352800), or vice versa,
while switching to another clock with the same base clock doesn't need
the (forcible) interface setup.

This patch implements the optimization for the TEAC clock quirk to
avoid the unnecessary interface re-setup.

Fixes: 5ce0b06a ("ALSA: usb-audio: Workaround for clock setup on TEAC devices")
Reported-by: Maris Abele <maris7abele@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20220531130749.30357-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

Merge series from Adam Thomson <DLG-Adam.Thomson.Opensource@dm.renesas.com>:

This series contains 2 small fixes around the AAD part of DA7219, particularly
in relation to jack pole detection on certain active headsets, and tidy up
when a jack is removed.

To avoid the unlikely possibility of register misalignment for
headphones being ungrounded/driven after a jack has been removed,
move the cancel_work_sync() call to the start of the jack removal
handling in the IRQ thread.
Signed-off-by: Adam Thomson <DLG-Adam.Thomson.Opensource@dm.renesas.com>
Link: https://lore.kernel.org/r/b3f9a679f1e27a9359dcecb496953c4af30acbaa.1653916368.git.DLG-Adam.Thomson.Opensource@dm.renesas.comSigned-off-by: Mark Brown <broonie@kernel.org>

It has been recently found that certain 'active' headsets can
be mis-detected as OMTP instead of CTIA, causing obvious issus
with audio quality. This relates to increased resistances which
negatively impacts the pole detection circuitry within the device.

To counter this, ground switches on both headphone channels are
available to enable/disable and these allow for the detection
process to operate as intended, even with active headsets. This
commit adds control of the ground switches to the AAD logic.
Signed-off-by: Adam Thomson <DLG-Adam.Thomson.Opensource@dm.renesas.com>
Link: https://lore.kernel.org/r/e0a627725c189dd50d6ce24571aed87fe2597395.1653916368.git.DLG-Adam.Thomson.Opensource@dm.renesas.comSigned-off-by: Mark Brown <broonie@kernel.org>

On some platforms (i.e. arc, m68k and sparc) __fls returns an int (while
on most platforms it returns an unsigned long). This triggers a format
warning on these few platforms as the driver uses %ld to print a warning.

So explicitly cast the return value to unsigned long to make the warning
go away (and so fix allmodconfig build on the affected architectures).

Fixes: beed9836 ("ASoC: Intel: avs: Machine board registration")
Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20220529141250.1979827-1-u.kleine-koenig@pengutronix.deSigned-off-by: Mark Brown <broonie@kernel.org>

Set microphone pins 0x18 (rear) and 0x19 (front) to VREF_50 to fix the
microphone noise on ASUS TUF B550M-PLUS which uses the ALCS1200A codec.
The initial value was VREF_80.

The same issue is also present on Windows using both the default Windows
driver and all tested Realtek drivers before version 6.0.9049.1. Comparing
Realtek driver 6.0.9049.1 (the first one without the microphone noise) to
Realtek driver 6.0.9047.1 (the last one with the microphone noise)
revealed that the fix is the result of setting pins 0x18 and 0x19 to
VREF_50.

This fix may also work for other boards that have been reported to have
the same microphone issue and use the ALC1150 and ALCS1200A codecs, since
these codecs are similar and the fix in the Realtek driver on Windows is
common for both. However, it is currently enabled only for ASUS TUF
B550M-PLUS as this is the only board that could be tested.
Signed-off-by: Marios Levogiannis <marios.levogiannis@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20220530074131.12258-1-marios.levogiannis@gmail.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

Void function return statements are not generally useful.
Signed-off-by: Xiang wangx <wangxiang@cdjrlc.com>
Link: https://lore.kernel.org/r/20220527121059.25221-1-wangxiang@cdjrlc.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

The 2022-model XPS 15 appears to use the same 4-speakers-on-ALC289
audio setup as the Dell XPS 15 9510, so requires the same quirk to
enable woofer output. Tested on my own 9520.

[ Move the entry to the right position in the SSID order -- tiwai ]

BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=216035
Cc: <stable@vger.kernel.org>
Signed-off-by: Rik van der Kemp <rik@upto11.nl>
Link: https://lore.kernel.org/r/181056a137b.d14baf90133058.8425453735588429828@upto11.nlSigned-off-by: Takashi Iwai <tiwai@suse.de>

Correct typo form sof-adl-mx98360a-nau8825.tplg to
sof-adl-max98360a-nau8825.tplg. The reason is tplg naming without naming
limitaion of length. It will be consistency with sof topology generation.
Signed-off-by: David Lin <CTLIN0@nuvoton.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20220526121301.1819541-1-CTLIN0@nuvoton.comSigned-off-by: Mark Brown <broonie@kernel.org>

At closing a USB MIDI output substream, there might be still a pending
work, which would eventually access the rawmidi runtime object that is
being released.  For fixing the race, make sure to cancel the pending
work at closing.

Reported-by: syzbot+6912c9592caca7ca0e7d@syzkaller.appspotmail.com
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/000000000000e7e75005dfd07cf6@google.com
Link: https://lore.kernel.org/r/20220525131203.11299-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

This patch implements a static mapping for Gigabyte B450/550 Mobos so
that the mixer elements appear reasonably and jack detections work
properly.
Reported-and-tested-by: Brock Szuszczewicz <brock@system76.com>
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=215988
Link: https://lore.kernel.org/r/20220525122018.3299-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

The pin "Platform Clock" was only used by the Intel Byt CR platform. In the
others, the error log will be informed. The patch will set the flag to
avoid the pin "Platform Clock" manipulated by the other platforms.
Signed-off-by: Oder Chiou <oder_chiou@realtek.com>
Reported-by: Sameer Pujar <spujar@nvidia.com>
Link: https://lore.kernel.org/r/20220516103055.20003-1-oder_chiou@realtek.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add new type for ALC245.
Signed-off-by: Kailang Yang <kailang@realtek.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/cef26a7cd3d146eb96a3994ce79e34d2@realtek.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

Devices such as the TC-Helicon GoXLR require the sync endpoint to be
configured in advance of the data endpoint in order for sound output
to work.

This patch simply changes the ordering of EP configuration to resolve
this.

Fixes: bf6313a0 ("ALSA: usb-audio: Refactor endpoint management")
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=215079Signed-off-by: Craig McLure <craig@mclure.net>
Reviewed-by: Jaroslav Kysela <perex@perex.cz>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20220524062115.25968-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

Add linux/module.h in acp-pci.c to solve the below dependency

All error/warnings (new ones prefixed by >>):

>> sound/soc/amd/acp/acp-pci.c:148:1: warning: data definition has no type or storage class
148 | MODULE_DEVICE_TABLE(pci, acp_pci_ids);
| ^~~~~~~~~~~~~~~~~~~
>> sound/soc/amd/acp/acp-pci.c:148:1: error: type defaults to 'int' in declaration of 'MODULE_DEVICE_TABLE' [-Werror=implicit-int]
...
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Syed Saba Kareem<ssabakar@amd.com>
Link: https://lore.kernel.org/r/20220523112956.3087604-1-ssabakar@amd.comSigned-off-by: Mark Brown <broonie@kernel.org>

Merge tag 'asoc-v5.19' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Updates for v5.19

This is quite a big update, partly due to the addition of some larger
drivers (more of which is to follow since at least the AVS driver is
still a work in progress) and partly due to Charles' work sorting out
our handling of endianness.  As has been the case recently it's much
more about drivers than the core.

 - Overhaul of endianness specification for data formats, avoiding
   needless restrictions due to CODECs.
 - Initial stages of Intel AVS driver merge.
 - Introduction of v4 IPC mechanism for SOF.
 - TDM mode support for AK4613.
 - Support for Analog Devices ADAU1361, Cirrus Logic CS35L45, Maxim
   MAX98396, MediaTek MT8186, NXP i.MX8 micfil and SAI interfaces,
   nVidia Tegra186 ASRC, and Texas Instruments TAS2764 and TAS2780

There are multiple xDR and xFR registers, the index is
from 0 to 7. FSL_SAI_xDR and FSL_SAI_xFR is abandoned,
replace them with FSL_SAI_xDR0 and FSL_SAI_xFR0.

Fixes: 4f7a0728 ("ASoC: fsl_sai: Add support for SAI new version")
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Link: https://lore.kernel.org/r/1653284661-18964-1-git-send-email-shengjiu.wang@nxp.comSigned-off-by: Mark Brown <broonie@kernel.org>

A obvious editing mistake caught with a cppcheck warning

sound/soc/soc-pcm.c:2132:8: style: Variable 'ret' is reassigned a
value before the old one has been used. [redundantAssignment]
   ret = soc_pcm_trigger(be_substream, cmd);
       ^
sound/soc/soc-pcm.c:2126:9: note: ret is assigned
    ret = soc_pcm_trigger(be_substream,
        ^
sound/soc/soc-pcm.c:2129:9: note: ret is assigned
    ret = soc_pcm_trigger(be_substream,
        ^

Fixes: 374b50e2 ('ASoC: soc-pcm: improve BE transition for TRIGGER_START')
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20220520210615.607229-1-pierre-louis.bossart@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Spelling mistake (triple letters) in comment.
Detected with the help of Coccinelle.
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Link: https://lore.kernel.org/r/20220521111145.81697-44-Julia.Lawall@inria.frSigned-off-by: Takashi Iwai <tiwai@suse.de>

Spelling mistake (triple letters) in comment.
Detected with the help of Coccinelle.
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Link: https://lore.kernel.org/r/20220521111145.81697-1-Julia.Lawall@inria.frSigned-off-by: Takashi Iwai <tiwai@suse.de>

Adds the PCI ID for X-Fi cards sold under the Platnum and XtremeMusic names

Before: snd_ctxfi 0000:05:05.0: chip 20K1 model Unknown (1102:0021) is found
After: snd_ctxfi 0000:05:05.0: chip 20K1 model SB046x (1102:0021) is found

[ This is only about defining the model name string, and the rest is
  handled just like before, as a default unknown device. 
  Edward confirmed that the stuff has been working fine -- tiwai ]
Signed-off-by: Edward Matijevic <motolav@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/cae7d1a4-8bd9-7dfe-7427-db7e766f7272@gmail.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

Merge for 5.18-rc1
Signed-off-by: Takashi Iwai <tiwai@suse.de>

The quirk entry for Focusrite Saffire 6 had no proper ep_idx for the
capture endpoint, and this confused the driver, resulting in the
broken sound.  This patch adds the missing ep_idx in the entry.

While we are at it, a couple of other entries (for Digidesign MBox and
MOTU MicroBook II) seem to have the same problem, and those are
covered as well.

Fixes: bf6313a0 ("ALSA: usb-audio: Refactor endpoint management")
Reported-by: André Kapelrud <a.kapelrud@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20220521065325.426-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

Maris reported that TEAC UD-501 (0644:8043) doesn't work with the
typical "clock source 41 is not valid, cannot use" errors on the
recent kernels.  The currently known workaround so far is to restore
(partially) what we've done unconditionally at the clock setup;
namely, re-setup the USB interface immediately after the clock is
changed.  This patch re-introduces the behavior conditionally for TEAC
devices.

Further notes:
- The USB interface shall be set later in
  snd_usb_endpoint_configure(), but this seems to be too late.
- Even calling  usb_set_interface() right after
  sne_usb_init_sample_rate() doesn't help; so this must be related
  with the clock validation, too.
- The device may still spew the "clock source 41 is not valid" error
  at the first clock setup.  This seems happening at the very first
  try of clock setup, but it disappears at later attempts.
  The error is likely harmless because the driver retries the clock
  setup (such an error is more or less expected on some devices).

Fixes: bf6313a0 ("ALSA: usb-audio: Refactor endpoint management")
Reported-and-tested-by: Maris Abele <maris7abele@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20220521064627.29292-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

GCC 12 sees that it's technically possible for num_streams to be larger
than ARRAY_SIZE(pcm->streams). Bounds-check the iterator.

../sound/pci/lola/lola_pcm.c: In function 'lola_pcm_update':
../sound/pci/lola/lola_pcm.c:567:64: warning: array subscript [0, 31] is outside array bounds of 'struct lola_stream[16]' [-Warray-bounds]
  567 |                         struct lola_stream *str = &pcm->streams[i];
      |                                                    ~~~~~~~~~~~~^~~
In file included from ../sound/pci/lola/lola_pcm.c:15:
../sound/pci/lola/lola.h:307:28: note: while referencing 'streams'
  307 |         struct lola_stream streams[MAX_STREAM_COUNT];
      |                            ^~~~~~~
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220520165537.2139826-1-keescook@chromium.orgSigned-off-by: Takashi Iwai <tiwai@suse.de>

Merge tag 'asoc-fix-v5.18-rc7' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Fix for v5.17

This is rather late and at this point I'm expecting it to get merged in
the merge window rather than as a fix but if we get a -rc8 it's a small,
driver specific fix which should be fine to send.

Merge series from Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
From the kernel point of view there are only few ops that needs to be exposed:

Hi,

SOF is using dma-trace (or dtrace) as a firmware tracing method, which is only
supported with IPC3 and it is not applicable for IPC4.

Currently the dtrace is 'open managed' regardless of IPC version (we do force
disable it for IPC4, but the dtrace calls remain in place).

From the kernel point of view there are only few ops that needs to be exposed
by the firmware tracing support and everything else is IPC private, should not
be known by the core.

This series converts the current dma-trace as ipc3 specific firmware tracing
sub-component and moves all private data out from generic code.

Regards,
Peter
---
Peter Ujfalusi (8):
  ASoC: SOF: Introduce IPC independent ops for firmware tracing support
  ASoC: SOF: Rename dtrace_is_supported flag to fw_trace_is_supported
  ASoC: SOF: Clone the trace code to ipc3-dtrace as fw_tracing
    implementation
  ASoC: SOF: Switch to IPC generic firmware tracing
  ASoC: SOF: ipc3-dtrace: Move host ops wrappers from generic header to
    private
  ASoC: SOF: Modify the host trace_init parameter list to include dmab
  ASoC: SOF: Introduce opaque storage of private data for firmware
    tracing
  ASoC: SOF: ipc3-dtrace: Move dtrace related variables local from
    sof_dev

 sound/soc/sof/Makefile          |   1 +
 sound/soc/sof/amd/acp-trace.c   |   4 +-
 sound/soc/sof/amd/acp.h         |   2 +-
 sound/soc/sof/core.c            |  13 +-
 sound/soc/sof/debug.c           |   2 +-
 sound/soc/sof/intel/hda-dsp.c   |   2 +-
 sound/soc/sof/intel/hda-trace.c |   4 +-
 sound/soc/sof/intel/hda.h       |   2 +-
 sound/soc/sof/ipc.c             |   6 +
 sound/soc/sof/ipc3-dtrace.c     | 649 ++++++++++++++++++++++++++++++++
 sound/soc/sof/ipc3-priv.h       |  38 ++
 sound/soc/sof/ipc3.c            |   3 +-
 sound/soc/sof/ops.c             |   2 +-
 sound/soc/sof/ops.h             |  26 --
 sound/soc/sof/pm.c              |   8 +-
 sound/soc/sof/sof-priv.h        |  53 +--
 sound/soc/sof/trace.c           | 621 ++----------------------------
 17 files changed, 767 insertions(+), 669 deletions(-)
 create mode 100644 sound/soc/sof/ipc3-dtrace.c

--
2.36.1

Validation of signed input should be done before casting to unsigned int.

Found by Linux Verification Center (linuxtesting.org) with SVACE.
Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Suggested-by: Mark Brown <broonie@kernel.org>
Fixes: 2fbe467b ("ASoC: max98090: Reject invalid values in custom control put()")
Link: https://lore.kernel.org/r/1652999486-29653-1-git-send-email-khoroshilov@ispras.ruSigned-off-by: Mark Brown <broonie@kernel.org>

The driver missed the default value of register 0xc320.
This patch adds that default value to avoid the error messages
when the driver went to suspend mode already.

BugLink: https://github.com/thesofproject/linux/issues/3651Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20220520090205.25857-1-shumingf@realtek.comSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

Add the ability to generate debug dumps on MediaTek SOF implementations.

Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

Remove two dependencies - issues reported by Intel kernel test bot.

From the datasheet, the only way to meet the lowest power consumption is
to pull low the 'pwdnn' gpio. But if it is low, IC will keep in
reset state, all registers reset to default.

And the power consumption is listed below
1. amp off and 'pwdnn' high => idle state, PVDD = 1mA, DVDD = 7mA
2. amp off and 'pwdnn' low  => shutdown state, PVDD < 20uA, DVDD < 15uA

It's the large difference for the consumption current

This fix is to use pm_runtime and regcache to handle 'pwdnn' gpio
control.
Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Link: https://lore.kernel.org/r/1652926418-8519-3-git-send-email-u0084500@gmail.comSigned-off-by: Mark Brown <broonie@kernel.org>

For RG 3byte read, the value order is  offset [0], [1], and [2].
Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Link: https://lore.kernel.org/r/1652926418-8519-2-git-send-email-u0084500@gmail.comSigned-off-by: Mark Brown <broonie@kernel.org>

RT1019 codec has two ways of controlling the en_spkr.  one way is
controlling through gpio pin method the another way is through codec
register update through driver.

Now Speaker enable/disable is controlled  through codec register updated
by codec driver.  This patch reverts gpio logic.

This reverts commit 5c5f08f7 ("ASoC:
amd: acp: Power on/off the speaker enable gpio pin based on DAPM
callback.")
Signed-off-by: V sujith kumar Reddy <Vsujithkumar.Reddy@amd.com>
Link: https://lore.kernel.org/r/20220516160619.17832-2-Vsujithkumar.Reddy@amd.comSigned-off-by: Mark Brown <broonie@kernel.org>

RT1019 codec has two ways of controlling the en_spkr.  one way is
controlling through gpio pin method the another way is through codec
register update through driver.

Now Speaker enable/disable is controlled  through codec register updated
by codec driver.  This patch reverts gpio logic.

This reverts commit 7fa5c33d ("ASoC:
amd: acp: Set gpio_spkr_en to None for max speaker amplifer in machine
driver").
Signed-off-by: V sujith kumar Reddy <Vsujithkumar.Reddy@amd.com>
Link: https://lore.kernel.org/r/20220516160619.17832-1-Vsujithkumar.Reddy@amd.comSigned-off-by: Mark Brown <broonie@kernel.org>

Fix the missing clk_disable_unprepare() before return
from wm2000_anc_transition() in the error handling case.

Fixes: 514cfd6d ("ASoC: wm2000: Integrate with clock API")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220514091053.686416-1-yangyingliang@huawei.comSigned-off-by: Mark Brown <broonie@kernel.org>

sound/soc/codecs/lpass-macro-common.c:28 lpass_macro_pds_init() warn: passing zero to 'PTR_ERR'
sound/soc/codecs/lpass-macro-common.c:38 lpass_macro_pds_init() warn: passing zero to 'PTR_ERR'
sound/soc/codecs/lpass-macro-common.c:54 lpass_macro_pds_init() warn: passing zero to 'ERR_PTR'

dev_pm_domain_attach_by_name() may return NULL, set 'ret' as
-ENODATA to fix this warning.

Fixes: 1a8ee4cf ("ASoC: codecs: Fix error handling in power domain init and exit handlers")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Link: https://lore.kernel.org/r/20220516120909.36356-1-yuehaibing@huawei.comSigned-off-by: Mark Brown <broonie@kernel.org>

Use pm_runtime_resume_and_get() to replace the pm_runtime_get_sync() and
pm_runtime_put_noidle() pattern.

No functional changes.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20220517173715.468894-1-pierre-louis.bossart@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>