linux-kernel - Re: [PATCH 5/5] ASoC: mediatek: mt8188-mt6359: Use bitfield macros for registers

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
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <0493414323dba8d5f59bca8cb7db53f300769dfa.camel@mediatek.com>
Date:   Thu, 8 Jun 2023 11:02:02 +0000
From:   Trevor Wu (吳文良) <Trevor.Wu@...iatek.com>
To:     "angelogioacchino.delregno@...labora.com" 
        <angelogioacchino.delregno@...labora.com>,
        "broonie@...nel.org" <broonie@...nel.org>
CC:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-mediatek@...ts.infradead.org" 
        <linux-mediatek@...ts.infradead.org>,
        "kernel@...labora.com" <kernel@...labora.com>,
        "tiwai@...e.com" <tiwai@...e.com>,
        "amergnat@...libre.com" <amergnat@...libre.com>,
        "lgirdwood@...il.com" <lgirdwood@...il.com>,
        "linux-arm-kernel@...ts.infradead.org" 
        <linux-arm-kernel@...ts.infradead.org>,
        "matthias.bgg@...il.com" <matthias.bgg@...il.com>,
        "perex@...ex.cz" <perex@...ex.cz>,
        "alsa-devel@...a-project.org" <alsa-devel@...a-project.org>,
        "dan.carpenter@...aro.org" <dan.carpenter@...aro.org>
Subject: Re: [PATCH 5/5] ASoC: mediatek: mt8188-mt6359: Use bitfield macros
 for registers

On Thu, 2023-06-08 at 12:07 +0200, AngeloGioacchino Del Regno wrote:
>  	 
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
>  Il 08/06/23 12:03, Trevor Wu (吳文良) ha scritto:
> > On Thu, 2023-06-08 at 10:47 +0200, 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_CFG0x032c
> >> + #define RG_TEST_ONBIT(0)
> >> + #define RG_TEST_TYPEBIT(2)
> >> +#define CKSYS_AUD_TOP_MON0x0330
> >> + #define TEST_MISO_COUNT_1GENMASK(3, 0)
> >> + #define TEST_MISO_COUNT_2GENMASK(7, 4)
> >> + #define TEST_MISO_DONE_1BIT(28)
> >> + #define TEST_MISO_DONE_2BIT(29)
> >> +
> >>   #define NAU8825_HS_PRESENTBIT(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;
> >>   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);
> > 
> > Hi Angelo,
> > 
> > Because CKSYS_AUD_TOP_CFG is a 32bit register, it should be better
> to
> > use regmap_set_bits instead.
> > 
> > regmap_set_bits(afe_priv->topckgen, CKSYS_AUD_TOP_CFG,
> RG_TEST_TYPE);
> > 
> 
> The previous call to regmap_update_bits() was unsetting RG_TEST_ON
> and RE_SW_RESET
> while setting RG_TEST_TYPE: using regmap_write() ensures that we do
> exactly the
> same, without the overhead of performing the additional register read
> and bit
> swapping operations.
> 
> Using the proposed regmap_set_bits() would change the behavior of
> this flow, which
> may result in unexpected hardware behavior, as we wouldn't be
> unsetting the
> previously mentioned two bits.
> 
> Regards,
> Angelo


It's unclear why the original author chose a mask value of 0xffff, but
remap_write() also clears the higher 16 bits which differs from the
original logic.

The comment for that line states 'set test type to synchronizer pulse',
so I suggest updating only BIT2 here. Additionally, I checked datasheet
for other two bits, which have default values of 0.

However, I came across an internal codebase that used regmap_write(),
so it should be a safe implementation. If you think regmap_write() is
better, it's ok with me.

Thanks,
Trevor

> 
> > Thanks,
> > Trevor
> > 
> > 
> >>    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)
> >> -- 
> >> 2.40.1
> 
> 
>