Re: [PATCH 1/3] mmc: sdhci-pci-o2micro: Add hardware tuning for eMMC

From: Craig Bergstrom
Date: Wed Jan 10 2018 - 19:53:16 EST

Next message: Sergey Senozhatsky: "[PATCH] workqueue: avoid hard lockups in show_workqueue_state()"
Previous message: Masami Hiramatsu: "[PATCH bpf-next v4 5/5] error-injection: Support fault injection framework"
In reply to: Shirley Her (SC): "RE: [PATCH 1/3] mmc: sdhci-pci-o2micro: Add hardware tuning for eMMC"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Gotcha.

Thank you for the detailed explanation. I will give that a try and see
if it fixes the problem.

Cheers,
Craig

On Wed, Jan 10, 2018 at 3:05 PM, Shirley Her (SC)
<shirley.her@xxxxxxxxxxxxxx> wrote:
> Hi, Craig:
>
> The upstream user is using SD card not eMMC for this device in the current market. BIOS does not need to set bit 4 of 0x308 to 1 for device that only use SD card. Since Bayhub already shipped this device in the market to support SD card and we can not just upstream driver to only support eMMC for this device.
>
> We can implement a driver that can detect SD or eMMC is used in the device but it will require a lot of code changes and retry which will delay the card init time plus there is no device+eMMC configuration in the market now.
>
> Since this is the 1st time this device is used to support eMMC, so we use 0x308 bit 4 to identify the device is connected to eMMC and BIOS can help to do the job which makes driver implementation simpler.
>
>
> Thanks,
> Shirley
>
>
> -----Original Message-----
> From: Craig Bergstrom [mailto:craigb@xxxxxxxxxx]
> Sent: Wednesday, January 10, 2018 2:33 PM
> To: Shirley Her (SC)
> Cc: Adrian Hunter; LinuxPatchCommit; ulf.hansson@xxxxxxxxxx; linux-mmc@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH 1/3] mmc: sdhci-pci-o2micro: Add hardware tuning for eMMC
>
> Hey Shirley,
>
> If all upstream users of this device need to set bit 4 of 0x308 to 1, should the upstream driver do that?
>
> Craig
>
>
> On Wed, Jan 10, 2018 at 12:05 PM, Shirley Her (SC) <shirley.her@xxxxxxxxxxxxxx> wrote:
>> Hi, Craig:
>>
>> As the message said during sign off as below. BIOS needs to set x0308 bit 4 to 1 so the driver will skip 3.3V in eMMC initialization process. It means depthcharge needs to set 0x308 bit 4 to 1 first in order for driver to use eMMC 1.8V.
>>
>> In some case of eMMC used as boot device, the eMMC signaling voltage
>>> is fixed to 1.8v, bios can set o2 sd host controller register 0x308
>>> bit4 to let host controller skip try 3.3.v signaling voltage in eMMC
>>> initialize process.
>>
>> Thanks,
>> Shirley
>>
>> -----Original Message-----
>> From: Craig Bergstrom [mailto:craigb@xxxxxxxxxx]
>> Sent: Thursday, December 21, 2017 9:18 AM
>> To: Adrian Hunter
>> Cc: LinuxPatchCommit; ulf.hansson@xxxxxxxxxx;
>> linux-mmc@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; Shirley Her
>> (SC)
>> Subject: Re: [PATCH 1/3] mmc: sdhci-pci-o2micro: Add hardware tuning
>> for eMMC
>>
>> Thanks for sending these patches out, I would be very excited to see these devices supported by Linux.
>>
>> I'm trying them out with a SEABIRD device (pci device 1217:8620).
>> When I boot up and load the driver, it don't see any indication of the existence of block device and I see these messages repeatedly spammed to the kernel logs:
>>
>> [ 0.855163] mmc1: Unknown controller version (3). You may
>> experience problems.
>> [ 0.862519] mmc1: SDHCI controller on PCI [0000:01:00.0] using ADMA
>> [ 93.152163] mmc1: Timeout waiting for hardware cmd interrupt.
>> [ 93.157911] mmc1: sdhci: ============ SDHCI REGISTER DUMP ===========
>> [ 93.164347] mmc1: sdhci: Sys addr: 0x00000000 | Version: 0x00000603
>> [ 93.170782] mmc1: sdhci: Blk size: 0x00000000 | Blk cnt: 0x00000000
>> [ 93.177218] mmc1: sdhci: Argument: 0x00000000 | Trn mode: 0x00000000
>> [ 93.183652] mmc1: sdhci: Present: 0x01ff0000 | Host ctl: 0x00000001
>> [ 93.190088] mmc1: sdhci: Power: 0x0000000f | Blk gap: 0x00000000
>> [ 93.196523] mmc1: sdhci: Wake-up: 0x00000000 | Clock: 0x0000fa07
>> [ 93.202956] mmc1: sdhci: Timeout: 0x00000000 | Int stat: 0x00018000
>> [ 93.209392] mmc1: sdhci: Int enab: 0x00ff0083 | Sig enab: 0x00ff0083
>> [ 93.215826] mmc1: sdhci: AC12 err: 0x00000000 | Slot int: 0x00000001
>> [ 93.222260] mmc1: sdhci: Caps: 0x25fcc8bf | Caps_1: 0x00002077
>> [ 93.228694] mmc1: sdhci: Cmd: 0x0000371a | Max curr: 0x005800c8
>> [ 93.235127] mmc1: sdhci: Resp[0]: 0x00000000 | Resp[1]: 0x00000000
>> [ 93.241562] mmc1: sdhci: Resp[2]: 0x00000000 | Resp[3]: 0x00000000
>> [ 93.247994] mmc1: sdhci: Host ctl2: 0x00000000
>> [ 93.252435] mmc1: sdhci: ADMA Err: 0x00000000 | ADMA Ptr: 0x00000000
>> [ 93.258863] mmc1: sdhci: ============================================
>> [ 103.392164] mmc1: Timeout waiting for hardware cmd interrupt.
>> [ 103.397913] mmc1: sdhci: ============ SDHCI REGISTER DUMP =========== [ 103.404350] mmc1: sdhci: Sys addr: 0x00000000 | Version: 0x00000603 [ 103.410786] mmc1: sdhci: Blk size: 0x00000000 | Blk cnt: 0x00000000 [ 103.417221] mmc1: sdhci: Argument: 0x00000000 | Trn mode: 0x00000000
>> [ 103.423654] mmc1: sdhci: Present: 0x01ff0000 | Host ctl: 0x00000001
>> [ 103.430089] mmc1: sdhci: Power: 0x0000000f | Blk gap: 0x00000000
>> [ 103.436523] mmc1: sdhci: Wake-up: 0x00000000 | Clock: 0x0000fa07
>> [ 103.442955] mmc1: sdhci: Timeout: 0x00000000 | Int stat: 0x00018000
>> [ 103.449390] mmc1: sdhci: Int enab: 0x00ff0083 | Sig enab: 0x00ff0083 [ 103.455823] mmc1: sdhci: AC12 err: 0x00000000 | Slot int: 0x00000001
>> [ 103.462255] mmc1: sdhci: Caps: 0x25fcc8bf | Caps_1: 0x00002077
>> [ 103.468689] mmc1: sdhci: Cmd: 0x0000371a | Max curr: 0x005800c8
>> [ 103.475126] mmc1: sdhci: Resp[0]: 0x00000000 | Resp[1]: 0x00000000
>> [ 103.481560] mmc1: sdhci: Resp[2]: 0x00000000 | Resp[3]: 0x00000000
>> [ 103.487994] mmc1: sdhci: Host ctl2: 0x00000000 [ 103.492435] mmc1:
>> sdhci: ADMA Err: 0x00000000 | ADMA Ptr: 0x00000000 [ 103.498867]
>> mmc1: sdhci: ============================================
>>
>> Is it expected that this should "just work", or am I doing something wrong? My patched kernel is based on 4.14.2.
>>
>> On Tue, Dec 19, 2017 at 2:05 AM, Adrian Hunter <adrian.hunter@xxxxxxxxx> wrote:
>>> On 04/12/17 11:39, LinuxPatchCommit wrote:
>>>> Dear All,
>>>>
>>>> For O2micro/Bayhubtech SD Host DeviceID 8620, eMMC HS200 mode is working at 1.8v and it uses hardware tuning. The hardware tuning only needs to send one tuning command instead of multiple tuning commands with software tuning.
>>>>
>>>> Signed-off-by: ernest.zhang <ernest.zhang@xxxxxxxxxxxxxx>
>>>> ---
>>>> drivers/mmc/host/sdhci-pci-o2micro.c | 276
>>>> ++++++++++++++++++++++++++++++-----
>>>> 1 file changed, 239 insertions(+), 37 deletions(-)
>>>>
>>>> diff --git a/drivers/mmc/host/sdhci-pci-o2micro.c
>>>> b/drivers/mmc/host/sdhci-pci-o2micro.c
>>>> index 14273ca00641..2a7ffd240497 100644
>>>> --- a/drivers/mmc/host/sdhci-pci-o2micro.c
>>>> +++ b/drivers/mmc/host/sdhci-pci-o2micro.c
>>>> @@ -16,22 +16,223 @@
>>>> */
>>>>
>>>> #include <linux/pci.h>
>>>> -
>>>> +#include <linux/platform_device.h>
>>>> +#include <linux/regulator/fixed.h>
>>>> +#include <linux/regulator/machine.h> #include <linux/mmc/host.h>
>>>> +#include <linux/mmc/mmc.h> #include <linux/delay.h>
>>>> #include "sdhci.h"
>>>> #include "sdhci-pci.h"
>>>> #include "sdhci-pci-o2micro.h"
>>>>
>>>> +static void sdhci_o2_start_tuning(struct sdhci_host *host) {
>>>> + u16 ctrl;
>>>> +
>>>> + ctrl = sdhci_readw(host, SDHCI_HOST_CONTROL2);
>>>> + ctrl |= SDHCI_CTRL_EXEC_TUNING;
>>>> + if (host->quirks2 & SDHCI_QUIRK2_TUNING_WORK_AROUND)
>>>> + ctrl |= SDHCI_CTRL_TUNED_CLK;
>>>> + sdhci_writew(host, ctrl, SDHCI_HOST_CONTROL2);
>>>> +
>>>> + /*
>>>> + * As per the Host Controller spec v3.00, tuning command
>>>> + * generates Buffer Read Ready interrupt, so enable that.
>>>> + *
>>>> + * Note: The spec clearly says that when tuning sequence
>>>> + * is being performed, the controller does not generate
>>>> + * interrupts other than Buffer Read Ready interrupt. But
>>>> + * to make sure we don't hit a controller bug, we _only_
>>>> + * enable Buffer Read Ready interrupt here.
>>>> + */
>>>> + sdhci_writel(host, SDHCI_INT_DATA_AVAIL, SDHCI_INT_ENABLE);
>>>> + sdhci_writel(host, SDHCI_INT_DATA_AVAIL, SDHCI_SIGNAL_ENABLE);
>>>> + }
>>>> +
>>>> +static void sdhci_o2_end_tuning(struct sdhci_host *host) {
>>>> + sdhci_writel(host, host->ier, SDHCI_INT_ENABLE);
>>>> + sdhci_writel(host, host->ier, SDHCI_SIGNAL_ENABLE); }
>>>> +
>>>> +static inline bool sdhci_data_line_cmd(struct mmc_command *cmd) {
>>>> + return cmd->data || cmd->flags & MMC_RSP_BUSY; }
>>>> +
>>>> +static void sdhci_del_timer(struct sdhci_host *host, struct
>>>> +mmc_request
>>>> +*mrq) {
>>>> + if (sdhci_data_line_cmd(mrq->cmd))
>>>> + del_timer(&host->data_timer);
>>>> + else
>>>> + del_timer(&host->timer); }
>>>> +
>>>> +static void sdhci_o2_set_tuning_mode(struct sdhci_host *host, bool
>>>> +hw) {
>>>> + u16 reg;
>>>> +
>>>> + if (hw) {
>>>> + // enable hardware tuning
>>>> + reg = sdhci_readw(host, O2_SD_VENDOR_SETTING);
>>>> + reg &= (~O2_SD_HW_TUNING_ENABLE);
>>>> + sdhci_writew(host, reg, O2_SD_VENDOR_SETTING);
>>>> + } else {
>>>> + reg = sdhci_readw(host, O2_SD_VENDOR_SETTING);
>>>> + reg |= O2_SD_HW_TUNING_ENABLE;
>>>> + sdhci_writew(host, reg, O2_SD_VENDOR_SETTING);
>>>> + }
>>>> +}
>>>> +
>>>> +static u8 data_buf[64];
>>>> +
>>>> +static void sdhci_o2_send_tuning(struct sdhci_host *host, u32 opcode) {
>>>> + struct mmc_command cmd = { };
>>>> + struct mmc_data data = { };
>>>> + struct scatterlist sg;
>>>> + struct mmc_request mrq = { };
>>>> + unsigned long flags;
>>>> + u32 b = host->sdma_boundary;
>>>> + int size = sizeof(data_buf);
>>>> +
>>>> + cmd.opcode = opcode;
>>>> + cmd.flags = MMC_RSP_PRESENT | MMC_RSP_OPCODE | MMC_RSP_CRC;
>>>> + cmd.mrq = &mrq;
>>>> + mrq.cmd = &cmd;
>>>> + mrq.data = &data;
>>>> + data.blksz = size;
>>>> + data.blocks = 1;
>>>> + data.flags = MMC_DATA_READ;
>>>> +
>>>> + data.timeout_ns = 150 * NSEC_PER_MSEC;
>>>> +
>>>> + data.sg = &sg;
>>>> + data.sg_len = 1;
>>>> + sg_init_one(&sg, data_buf, size);
>>>> +
>>>> + spin_lock_irqsave(&host->lock, flags);
>>>> +
>>>> + sdhci_writew(host, SDHCI_MAKE_BLKSZ(b, 64), SDHCI_BLOCK_SIZE);
>>>> +
>>>> + /*
>>>> + * The tuning block is sent by the card to the host controller.
>>>> + * So we set the TRNS_READ bit in the Transfer Mode register.
>>>> + * This also takes care of setting DMA Enable and Multi Block
>>>> + * Select in the same register to 0.
>>>> + */
>>>> + sdhci_writew(host, SDHCI_TRNS_READ, SDHCI_TRANSFER_MODE);
>>>> +
>>>> + sdhci_send_command(host, &cmd);
>>>> +
>>>> + host->cmd = NULL;
>>>> +
>>>> + sdhci_del_timer(host, &mrq);
>>>> +
>>>> + host->tuning_done = 0;
>>>> +
>>>> + mmiowb();
>>>> + spin_unlock_irqrestore(&host->lock, flags);
>>>> +
>>>> + /* Wait for Buffer Read Ready interrupt */
>>>> + wait_event_timeout(host->buf_ready_int, (host->tuning_done == 1),
>>>> + msecs_to_jiffies(50));
>>>> +
>>>> +}
>>>> +
>>>> +static void sdhci_o2_reset_tuning(struct sdhci_host *host) {
>>>> + u16 ctrl;
>>>> +
>>>> + ctrl = sdhci_readw(host, SDHCI_HOST_CONTROL2);
>>>> + ctrl &= ~SDHCI_CTRL_TUNED_CLK;
>>>> + ctrl &= ~SDHCI_CTRL_EXEC_TUNING;
>>>> + sdhci_writew(host, ctrl, SDHCI_HOST_CONTROL2); }
>>>> +
>>>> +static void __sdhci_o2_execute_tuning(struct sdhci_host *host, u32
>>>> +opcode) {
>>>> +
>>>> + int i;
>>>> +
>>>> + sdhci_o2_send_tuning(host, MMC_SEND_TUNING_BLOCK_HS200);
>>>
>>> So it looks like you send the tuning command only once. Is that right?
>>> Maybe you could briefly outline what the tuning procedure is, and how
>>> it differs from sdhci_execute_tuning().
>>>
>>>> +
>>>> + for (i = 0; i < 150; i++) {
>>>> + u16 ctrl = sdhci_readw(host, SDHCI_HOST_CONTROL2);
>>>> +
>>>> + if (!(ctrl & SDHCI_CTRL_EXEC_TUNING)) {
>>>> + if (ctrl & SDHCI_CTRL_TUNED_CLK) {
>>>> + pr_info("%s: HW tuning ok !\n",
>>>> + mmc_hostname(host->mmc));
>>>> + host->tuning_done = true;
>>>> + return;
>>>> + }
>>>> + pr_warn("%s: HW tuning failed !\n",
>>>> + mmc_hostname(host->mmc));
>>>> +
>>>> + break;
>>>> + }
>>>> +
>>>> + mdelay(1);
>>>> + }
>>>> +
>>>> + pr_info("%s: Tuning failed, falling back to fixed sampling clock\n",
>>>> + mmc_hostname(host->mmc));
>>>> + sdhci_o2_reset_tuning(host);
>>>> +
>>>> +}
>>>> +
>>>> +static int sdhci_o2_execute_tuning(struct mmc_host *mmc, u32 opcode) {
>>>> + struct sdhci_host *host = mmc_priv(mmc);
>>>> + int current_bus_width = 0;
>>>> +
>>>> + /*
>>>> + * This handler only implements the eMMC tuning that is specific to
>>>> + * this controller. Fall back to the standard method for other TINMING.
>>>> + */
>>>> + if (host->timing != MMC_TIMING_MMC_HS200)
>>>> + return sdhci_execute_tuning(mmc, opcode);
>>>> +
>>>> + if (WARN_ON(opcode != MMC_SEND_TUNING_BLOCK_HS200))
>>>> + return -EINVAL;
>>>> +
>>>> + /*
>>>> + * o2 sdhci host didn't support 8bit emmc tuning
>>>> + */
>>>> + if (mmc->ios.bus_width == MMC_BUS_WIDTH_8) {
>>>> + current_bus_width = mmc->ios.bus_width;
>>>> + mmc->ios.bus_width = MMC_BUS_WIDTH_4;
>>>> + mmc->ops->set_ios(mmc, &mmc->ios);
>>>> + }
>>>> +
>>>> + sdhci_o2_set_tuning_mode(host, true);
>>>> +
>>>> + sdhci_o2_start_tuning(host);
>>>> +
>>>> + __sdhci_o2_execute_tuning(host, opcode);
>>>> +
>>>> + sdhci_o2_end_tuning(host);
>>>> +
>>>> + if (current_bus_width == MMC_BUS_WIDTH_8) {
>>>> + mmc->ios.bus_width = current_bus_width;
>>>> + mmc->ops->set_ios(mmc, &mmc->ios);
>>>> + }
>>>> +
>>>> + host->flags &= ~SDHCI_HS400_TUNING;
>>>> + return 0;
>>>> +}
>>>> +
>>>> static void o2_pci_set_baseclk(struct sdhci_pci_chip *chip, u32 value) {
>>>> u32 scratch_32;
>>>> - pci_read_config_dword(chip->pdev,
>>>> - O2_SD_PLL_SETTING, &scratch_32);
>>>> + pci_read_config_dword(chip->pdev, O2_SD_PLL_SETTING,
>>>> + &scratch_32);
>>>>
>>>> scratch_32 &= 0x0000FFFF;
>>>> scratch_32 |= value;
>>>>
>>>> - pci_write_config_dword(chip->pdev,
>>>> - O2_SD_PLL_SETTING, scratch_32);
>>>> + pci_write_config_dword(chip->pdev, O2_SD_PLL_SETTING,
>>>> + scratch_32);
>>>> }
>>>>
>>>> static void o2_pci_led_enable(struct sdhci_pci_chip *chip) @@ -40,23 +241,19 @@ static void o2_pci_led_enable(struct sdhci_pci_chip *chip)
>>>> u32 scratch_32;
>>>>
>>>> /* Set led of SD host function enable */
>>>> - ret = pci_read_config_dword(chip->pdev,
>>>> - O2_SD_FUNC_REG0, &scratch_32);
>>>> + ret = pci_read_config_dword(chip->pdev, O2_SD_FUNC_REG0,
>>>> + &scratch_32);
>>>> if (ret)
>>>> return;
>>>>
>>>> scratch_32 &= ~O2_SD_FREG0_LEDOFF;
>>>> - pci_write_config_dword(chip->pdev,
>>>> - O2_SD_FUNC_REG0, scratch_32);
>>>> + pci_write_config_dword(chip->pdev, O2_SD_FUNC_REG0,
>>>> + scratch_32);
>>>>
>>>> - ret = pci_read_config_dword(chip->pdev,
>>>> - O2_SD_TEST_REG, &scratch_32);
>>>> + ret = pci_read_config_dword(chip->pdev, O2_SD_TEST_REG,
>>>> + &scratch_32);
>>>> if (ret)
>>>> return;
>>>>
>>>> scratch_32 |= O2_SD_LED_ENABLE;
>>>> - pci_write_config_dword(chip->pdev,
>>>> - O2_SD_TEST_REG, scratch_32);
>>>> + pci_write_config_dword(chip->pdev, O2_SD_TEST_REG,
>>>> + scratch_32);
>>>>
>>>> }
>>>>
>>>> @@ -104,8 +301,7 @@ static void sdhci_pci_o2_fujin2_pci_init(struct sdhci_pci_chip *chip)
>>>> scratch_32 |= 0x00CC;
>>>> pci_write_config_dword(chip->pdev, O2_SD_CAP_REG0, scratch_32);
>>>> /* Set DLL Tuning Window */
>>>> - ret = pci_read_config_dword(chip->pdev,
>>>> - O2_SD_TUNING_CTRL, &scratch_32);
>>>> + ret = pci_read_config_dword(chip->pdev, O2_SD_TUNING_CTRL,
>>>> +&scratch_32);
>>>> if (ret)
>>>> return;
>>>> scratch_32 &= ~(0x000000FF);
>>>> @@ -137,8 +333,7 @@ static void sdhci_pci_o2_fujin2_pci_init(struct sdhci_pci_chip *chip)
>>>> scratch_32 |= 0x30000000;
>>>> pci_write_config_dword(chip->pdev, O2_SD_CAPS, scratch_32);
>>>>
>>>> - ret = pci_read_config_dword(chip->pdev,
>>>> - O2_SD_MISC_CTRL4, &scratch_32);
>>>> + ret = pci_read_config_dword(chip->pdev, O2_SD_MISC_CTRL4,
>>>> +&scratch_32);
>>>> if (ret)
>>>> return;
>>>> scratch_32 &= ~(0x000f0000);
>>>> @@ -151,6 +346,7 @@ int sdhci_pci_o2_probe_slot(struct sdhci_pci_slot *slot)
>>>> struct sdhci_pci_chip *chip;
>>>> struct sdhci_host *host;
>>>> u32 reg;
>>>> + int ret;
>>>>
>>>> chip = slot->chip;
>>>> host = slot->host;
>>>> @@ -164,6 +360,22 @@ int sdhci_pci_o2_probe_slot(struct sdhci_pci_slot *slot)
>>>> if (reg & 0x1)
>>>> host->quirks |=
>>>> SDHCI_QUIRK_MULTIBLOCK_READ_ACMD12;
>>>>
>>>> + if (chip->pdev->device == PCI_DEVICE_ID_O2_SEABIRD0) {
>>>> + ret = pci_read_config_dword(chip->pdev,
>>>> + O2_SD_MISC_SETTING, &reg);
>>>> + if (ret)
>>>> + return -EIO;
>>>> + if (reg & (1 << 4)) {
>>>> + pr_info("%s: emmc 1.8v flag is set,
>>>> + force 1.8v signaling voltage\n",
>>>> + mmc_hostname(host->mmc));
>>>> + host->flags &= ~(SDHCI_SIGNALING_330);
>>>> + host->flags |= SDHCI_SIGNALING_180;
>>>> + }
>>>> + }
>>>> +
>>>> + host->mmc_host_ops.execute_tuning =
>>>> + sdhci_o2_execute_tuning;
>>>> +
>>>> if (chip->pdev->device != PCI_DEVICE_ID_O2_FUJIN2)
>>>> break;
>>>> /* set dll watch dog timer */ @@ -191,8 +403,7 @@ int
>>>> sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> case PCI_DEVICE_ID_O2_8320:
>>>> case PCI_DEVICE_ID_O2_8321:
>>>> /* This extra setup is required due to broken ADMA. */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_LOCK_WP, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_LOCK_WP,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>> scratch &= 0x7f;
>>>> @@ -202,8 +413,7 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> pci_write_config_byte(chip->pdev, O2_SD_MULTI_VCC3V,
>>>> 0x08);
>>>>
>>>> /* Disable CLK_REQ# support after media DET */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_CLKREQ, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_CLKREQ,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>> scratch |= 0x20;
>>>> @@ -224,16 +434,14 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> pci_write_config_byte(chip->pdev, O2_SD_ADMA2, 0x08);
>>>>
>>>> /* Disable the infinite transfer mode */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_INF_MOD, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_INF_MOD,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>> scratch |= 0x08;
>>>> pci_write_config_byte(chip->pdev, O2_SD_INF_MOD,
>>>> scratch);
>>>>
>>>> /* Lock WP */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_LOCK_WP, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_LOCK_WP,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>> scratch |= 0x80;
>>>> @@ -243,8 +451,7 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> case PCI_DEVICE_ID_O2_SDS1:
>>>> case PCI_DEVICE_ID_O2_FUJIN2:
>>>> /* UnLock WP */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_LOCK_WP, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_LOCK_WP,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>>
>>>> @@ -319,15 +526,13 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> if (ret)
>>>> return ret;
>>>> scratch_32 &= ~(0xE0);
>>>> - pci_write_config_dword(chip->pdev,
>>>> - O2_SD_CAP_REG2, scratch_32);
>>>> + pci_write_config_dword(chip->pdev, O2_SD_CAP_REG2,
>>>> + scratch_32);
>>>>
>>>> if (chip->pdev->device == PCI_DEVICE_ID_O2_FUJIN2)
>>>> sdhci_pci_o2_fujin2_pci_init(chip);
>>>>
>>>> /* Lock WP */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_LOCK_WP, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_LOCK_WP,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>> scratch |= 0x80;
>>>> @@ -336,8 +541,7 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> case PCI_DEVICE_ID_O2_SEABIRD0:
>>>> case PCI_DEVICE_ID_O2_SEABIRD1:
>>>> /* UnLock WP */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_LOCK_WP, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_LOCK_WP,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>>
>>>> @@ -369,11 +573,9 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
>>>> }
>>>>
>>>> /* Set Tuning Windows to 5 */
>>>> - pci_write_config_byte(chip->pdev,
>>>> - O2_SD_TUNING_CTRL, 0x55);
>>>> + pci_write_config_byte(chip->pdev, O2_SD_TUNING_CTRL,
>>>> + 0x55);
>>>> /* Lock WP */
>>>> - ret = pci_read_config_byte(chip->pdev,
>>>> - O2_SD_LOCK_WP, &scratch);
>>>> + ret = pci_read_config_byte(chip->pdev, O2_SD_LOCK_WP,
>>>> + &scratch);
>>>> if (ret)
>>>> return ret;
>>>> scratch |= 0x80;
>>>> --
>>>> 2.14.1
>>>>
>>>

Next message: Sergey Senozhatsky: "[PATCH] workqueue: avoid hard lockups in show_workqueue_state()"
Previous message: Masami Hiramatsu: "[PATCH bpf-next v4 5/5] error-injection: Support fault injection framework"
In reply to: Shirley Her (SC): "RE: [PATCH 1/3] mmc: sdhci-pci-o2micro: Add hardware tuning for eMMC"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]