| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <a8b454e3-dded-02d6-b97a-6a7aa3420cf1@collabora.com>
Date: Thu, 8 Jun 2023 11:57:30 +0200
From: AngeloGioacchino Del Regno
<angelogioacchino.delregno@...labora.com>
To: Alexandre Mergnat <amergnat@...libre.com>, broonie@...nel.org
Cc: lgirdwood@...il.com, perex@...ex.cz, tiwai@...e.com,
matthias.bgg@...il.com, trevor.wu@...iatek.com,
dan.carpenter@...aro.org, alsa-devel@...a-project.org,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
linux-mediatek@...ts.infradead.org, kernel@...labora.com
Subject: Re: [PATCH 5/5] ASoC: mediatek: mt8188-mt6359: Use bitfield macros
for registers
Il 08/06/23 11:50, Alexandre Mergnat ha scritto:
> On 08/06/2023 10:47, AngeloGioacchino Del Regno wrote:
>> Replace open coded instances of FIELD_GET() with it, move register
>> definitions at the top of the file and also replace magic numbers
>> with register definitions.
>>
>> While at it, also change a regmap_update_bits() call to regmap_write()
>> because the top 29 bits of AUD_TOP_CFG (31:3) are reserved (unused).
>>
>> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>
>> ---
>> sound/soc/mediatek/mt8188/mt8188-mt6359.c | 32 ++++++++++++++---------
>> 1 file changed, 19 insertions(+), 13 deletions(-)
>>
>> diff --git a/sound/soc/mediatek/mt8188/mt8188-mt6359.c
>> b/sound/soc/mediatek/mt8188/mt8188-mt6359.c
>> index 5b2660139421..ac69c23e0da1 100644
>> --- a/sound/soc/mediatek/mt8188/mt8188-mt6359.c
>> +++ b/sound/soc/mediatek/mt8188/mt8188-mt6359.c
>> @@ -6,6 +6,7 @@
>> * Author: Trevor Wu <trevor.wu@...iatek.com>
>> */
>> +#include <linux/bitfield.h>
>> #include <linux/input.h>
>> #include <linux/module.h>
>> #include <linux/of_device.h>
>> @@ -19,6 +20,15 @@
>> #include "../common/mtk-afe-platform-driver.h"
>> #include "../common/mtk-soundcard-driver.h"
>> +#define CKSYS_AUD_TOP_CFG 0x032c
>> + #define RG_TEST_ON BIT(0)
>> + #define RG_TEST_TYPE BIT(2)
>> +#define CKSYS_AUD_TOP_MON 0x0330
>> + #define TEST_MISO_COUNT_1 GENMASK(3, 0)
>> + #define TEST_MISO_COUNT_2 GENMASK(7, 4)
>> + #define TEST_MISO_DONE_1 BIT(28)
>> + #define TEST_MISO_DONE_2 BIT(29)
>> +
>> #define NAU8825_HS_PRESENT BIT(0)
>> /*
>> @@ -251,9 +261,6 @@ static const struct snd_kcontrol_new
>> mt8188_nau8825_controls[] = {
>> SOC_DAPM_PIN_SWITCH("Headphone Jack"),
>> };
>> -#define CKSYS_AUD_TOP_CFG 0x032c
>> -#define CKSYS_AUD_TOP_MON 0x0330
>> -
>> static int mt8188_mt6359_mtkaif_calibration(struct snd_soc_pcm_runtime *rtd)
>> {
>> struct snd_soc_component *cmpnt_afe =
>> @@ -265,13 +272,13 @@ static int mt8188_mt6359_mtkaif_calibration(struct
>> snd_soc_pcm_runtime *rtd)
>> struct mtkaif_param *param;
>> int chosen_phase_1, chosen_phase_2;
>> int prev_cycle_1, prev_cycle_2;
>> - int test_done_1, test_done_2;
>> + u8 test_done_1, test_done_2;
>> int cycle_1, cycle_2;
>
> Shouldn't be unsigned too ?
>
> I'm wondering if it would be better (probably in another patch) to change the data
> type of the other variables too, to match their use-case. (maybe it's already the
> case, I'm just wondering)
>
In theory, yes, cycle_1 and 2 should be unsigned, but the signedness of this
variable is getting used in the calibration logic later, as in the for loop
they are both being reinitialized to -1 ... so I couldn't just switch 'em to
unsigned.
>> int mtkaif_chosen_phase[MT8188_MTKAIF_MISO_NUM];
>> int mtkaif_phase_cycle[MT8188_MTKAIF_MISO_NUM];
>> int mtkaif_calibration_num_phase;
>> bool mtkaif_calibration_ok;
>> - unsigned int monitor = 0;
>> + u32 monitor = 0;
>> int counter;
>> int phase;
>> int i;
>> @@ -303,8 +310,7 @@ static int mt8188_mt6359_mtkaif_calibration(struct
>> snd_soc_pcm_runtime *rtd)
>> mt6359_mtkaif_calibration_enable(cmpnt_codec);
>> /* set test type to synchronizer pulse */
>> - regmap_update_bits(afe_priv->topckgen,
>> - CKSYS_AUD_TOP_CFG, 0xffff, 0x4);
>> + regmap_write(afe_priv->topckgen, CKSYS_AUD_TOP_CFG, RG_TEST_TYPE);
>> mtkaif_calibration_num_phase = 42; /* mt6359: 0 ~ 42 */
>> mtkaif_calibration_ok = true;
>> @@ -314,7 +320,7 @@ static int mt8188_mt6359_mtkaif_calibration(struct
>> snd_soc_pcm_runtime *rtd)
>> mt6359_set_mtkaif_calibration_phase(cmpnt_codec,
>> phase, phase, phase);
>> - regmap_set_bits(afe_priv->topckgen, CKSYS_AUD_TOP_CFG, 0x1);
>> + regmap_set_bits(afe_priv->topckgen, CKSYS_AUD_TOP_CFG, RG_TEST_ON);
>> test_done_1 = 0;
>> test_done_2 = 0;
>> @@ -326,14 +332,14 @@ static int mt8188_mt6359_mtkaif_calibration(struct
>> snd_soc_pcm_runtime *rtd)
>> while (!(test_done_1 & test_done_2)) {
>> regmap_read(afe_priv->topckgen,
>> CKSYS_AUD_TOP_MON, &monitor);
>> - test_done_1 = (monitor >> 28) & 0x1;
>> - test_done_2 = (monitor >> 29) & 0x1;
>> + test_done_1 = FIELD_GET(TEST_MISO_DONE_1, monitor);
>> + test_done_2 = FIELD_GET(TEST_MISO_DONE_2, monitor);
>> if (test_done_1 == 1)
>> - cycle_1 = monitor & 0xf;
>> + cycle_1 = FIELD_GET(TEST_MISO_COUNT_1, monitor);
>> if (test_done_2 == 1)
>> - cycle_2 = (monitor >> 4) & 0xf;
>> + cycle_2 = FIELD_GET(TEST_MISO_COUNT_2, monitor);
>> /* handle if never test done */
>> if (++counter > 10000) {
>> @@ -361,7 +367,7 @@ static int mt8188_mt6359_mtkaif_calibration(struct
>> snd_soc_pcm_runtime *rtd)
>> mtkaif_phase_cycle[MT8188_MTKAIF_MISO_1] = prev_cycle_2;
>> }
>> - regmap_clear_bits(afe_priv->topckgen, CKSYS_AUD_TOP_CFG, 0x1);
>> + regmap_clear_bits(afe_priv->topckgen, CKSYS_AUD_TOP_CFG, RG_TEST_ON);
>> if (mtkaif_chosen_phase[MT8188_MTKAIF_MISO_0] >= 0 &&
>> mtkaif_chosen_phase[MT8188_MTKAIF_MISO_1] >= 0)
>
> After that:
>
> Reviewed-by: Alexandre Mergnat <amergnat@...libre.com>
>
Powered by blists - more mailing lists