| Message ID | 20201224172010.10701-1-huobean@gmail.com |
|---|---|
| Headers | show |
| Series | Two changes of UFS sysfs | expand |
On 2020-12-25 01:20, Bean Huo wrote: > From: Bean Huo <beanhuo@micron.com> > > The race issue may exist between UFS access in UFS sysfs context and > UFS > shutdown, thus will cause pm_runtime_get_sync() resume failure. Are you trying to fix the race condition by adding these checks or just adding these checks in case pm_runtime_get_sync() fails? Can Guo. > Add handling of the return value of pm_runtime_get_sync(). Let it > return > in case pm_runtime_get_sync() resume failed. > > Signed-off-by: Bean Huo <beanhuo@micron.com> > --- > drivers/scsi/ufs/ufs-sysfs.c | 38 ++++++++++++++++++++++++++++++------ > 1 file changed, 32 insertions(+), 6 deletions(-) > > diff --git a/drivers/scsi/ufs/ufs-sysfs.c > b/drivers/scsi/ufs/ufs-sysfs.c > index 0e1438485133..8e5e36e01bee 100644 > --- a/drivers/scsi/ufs/ufs-sysfs.c > +++ b/drivers/scsi/ufs/ufs-sysfs.c > @@ -154,12 +154,17 @@ static ssize_t auto_hibern8_show(struct device > *dev, > struct device_attribute *attr, char *buf) > { > u32 ahit; > + int ret; > struct ufs_hba *hba = dev_get_drvdata(dev); > > if (!ufshcd_is_auto_hibern8_supported(hba)) > return -EOPNOTSUPP; > > - pm_runtime_get_sync(hba->dev); > + ret = pm_runtime_get_sync(hba->dev); > + if (ret < 0) { > + pm_runtime_put_noidle(hba->dev); > + return ret; > + } > ufshcd_hold(hba, false); > ahit = ufshcd_readl(hba, REG_AUTO_HIBERNATE_IDLE_TIMER); > ufshcd_release(hba); > @@ -225,7 +230,12 @@ static ssize_t ufs_sysfs_read_desc_param(struct > ufs_hba *hba, > if (param_size > 8) > return -EINVAL; > > - pm_runtime_get_sync(hba->dev); > + ret = pm_runtime_get_sync(hba->dev); > + if (ret < 0) { > + pm_runtime_put_noidle(hba->dev); > + return ret; > + } > + > ret = ufshcd_read_desc_param(hba, desc_id, desc_index, > param_offset, desc_buf, param_size); > pm_runtime_put_sync(hba->dev); > @@ -594,7 +604,11 @@ static ssize_t _name##_show(struct device > *dev, \ > desc_buf = kzalloc(QUERY_DESC_MAX_SIZE, GFP_ATOMIC); \ > if (!desc_buf) \ > return -ENOMEM; \ > - pm_runtime_get_sync(hba->dev); \ > + ret = pm_runtime_get_sync(hba->dev); \ > + if (ret < 0) { \ > + pm_runtime_put_noidle(hba->dev); \ > + return ret; \ > + } \ > ret = ufshcd_query_descriptor_retry(hba, \ > UPIU_QUERY_OPCODE_READ_DESC, QUERY_DESC_IDN_DEVICE, \ > 0, 0, desc_buf, &desc_len); \ > @@ -653,7 +667,11 @@ static ssize_t _name##_show(struct device > *dev, \ > struct ufs_hba *hba = dev_get_drvdata(dev); \ > if (ufshcd_is_wb_flags(QUERY_FLAG_IDN##_uname)) \ > index = ufshcd_wb_get_query_index(hba); \ > - pm_runtime_get_sync(hba->dev); \ > + ret = pm_runtime_get_sync(hba->dev); \ > + if (ret < 0) { \ > + pm_runtime_put_noidle(hba->dev); \ > + return ret; \ > + } \ > ret = ufshcd_query_flag(hba, UPIU_QUERY_OPCODE_READ_FLAG, \ > QUERY_FLAG_IDN##_uname, index, &flag); \ > pm_runtime_put_sync(hba->dev); \ > @@ -711,7 +729,11 @@ static ssize_t _name##_show(struct device > *dev, \ > u8 index = 0; \ > if (ufshcd_is_wb_attrs(QUERY_ATTR_IDN##_uname)) \ > index = ufshcd_wb_get_query_index(hba); \ > - pm_runtime_get_sync(hba->dev); \ > + ret = pm_runtime_get_sync(hba->dev); \ > + if (ret < 0) { \ > + pm_runtime_put_noidle(hba->dev); \ > + return ret; \ > + } \ > ret = ufshcd_query_attr(hba, UPIU_QUERY_OPCODE_READ_ATTR, \ > QUERY_ATTR_IDN##_uname, index, 0, &value); \ > pm_runtime_put_sync(hba->dev); \ > @@ -850,7 +872,11 @@ static ssize_t > dyn_cap_needed_attribute_show(struct device *dev, > u8 lun = ufshcd_scsi_to_upiu_lun(sdev->lun); > int ret; > > - pm_runtime_get_sync(hba->dev); > + ret = pm_runtime_get_sync(hba->dev); > + if (ret < 0) { > + pm_runtime_put_noidle(hba->dev); > + return ret; > + } > ret = ufshcd_query_attr(hba, UPIU_QUERY_OPCODE_READ_ATTR, > QUERY_ATTR_IDN_DYN_CAP_NEEDED, lun, 0, &value); > pm_runtime_put_sync(hba->dev);
On Mon, 2020-12-28 at 09:50 +0800, Can Guo wrote: > > The race issue may exist between UFS access in UFS sysfs context > > and > > UFS > > shutdown, thus will cause pm_runtime_get_sync() resume failure. > > Are you trying to fix the race condition by adding these checks or > just > adding these checks in case pm_runtime_get_sync() fails? > > Can Guo. Can, thanks for your review. Sorry, I didn't quite get your point. This patch is just to add checkup in case pm_runtime_get_sync() failed. what else should be added? Bean
From: Bean Huo <beanhuo@micron.com> Changelog: V1---v2: 1. Add new patch "Let resume callback return -EBUSY after ufshcd_shutdown", Because the ufshcd_*_resume still returns successful result 0 after ufshcd_shutdown(). Even add handling of the return value of pm_runtime_get_sync(), but still there is timeout. Bean Huo (3): scsi: ufs: Replace sprintf and snprintf with sysfs_emit scsi: ufs: Add handling of the return value of pm_runtime_get_sync() scsi: ufs: Let resume callback return -EBUSY after ufshcd_shutdown drivers/scsi/ufs/ufs-sysfs.c | 68 +++++++++++++++++++++++++----------- drivers/scsi/ufs/ufshcd.c | 12 ++++--- 2 files changed, 55 insertions(+), 25 deletions(-)