phy: for 4.0 -rc cycle
Subsystems (3)
Hi Greg,
Please find the pull request for 4.0 -rc cycle. We should thank Axel Lin for reviewing all the PHY drivers and sending fixes. This also includes a fix from Thierry Reding in phy core.
Let me know if I have to change something.
Thanks
Kishon
The following changes since commit c517d838eb7d07bbe9507871fab3931deccff539:
Linux 4.0-rc1 (2015-02-22 18:21:14 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux-phy.git tags/for-4.0-rc
for you to fetch changes up to b1ff3231b2d4197ef5024f9c57ffc6cfa562590c:
phy: omap-usb2: Fix missing clk_prepare call when using old dt name (2015-03-13 17:14:39 +0530)
----------------------------------------------------------------
contains fixes all over drivers/phy
----------------------------------------------------------------
Axel Lin (19):
phy: miphy28lp: Avoid calling of_get_child_count() multiple times
phy: miphy365x: Avoid calling of_get_child_count() multiple times
phy: armada375-usb2: Set drvdata for phy and use it
phy: xgene: Remove duplicate code to set ctx->dev
phy: miphy28lp: Add missing .owner field in miphy28lp_ops
phy: exynos-mipi-video: Fixup the test for state->regmap
phy: exynos-mipi-video: Use spin_lock to protct state->regmap rmw operations
phy: exynos-dp-video: Kill exynos_dp_video_phy_pwr_isol function
phy: hix5hd2-sata: Check return value of platform_get_resource
phy: samsung-usb2: Remove NULL terminating entry from phys array
phy: ti-pipe3: Simplify ti_pipe3_dpll_wait_lock implementation
phy: rockchip-usb: Fixup rockchip_usb_phy_power_on failure path
phy: exynos5-usbdrd: Fix off-by-one valid value checking for args->args[0]
phy: twl4030-usb: Remove redundant assignment for twl->linkstat
phy: miphy365x: Convert to devm_kcalloc and fix wrong sizeof
phy: miphy28lp: Convert to devm_kcalloc and fix wrong sizof
phy: core: Fixup return value of phy_exit when !pm_runtime_enabled
phy: ti/omap: Fix modalias
phy: omap-usb2: Fix missing clk_prepare call when using old dt name
Thierry Reding (1):
phy: Find the right match in devm_phy_destroy()
drivers/phy/phy-armada375-usb2.c | 3 ++- drivers/phy/phy-core.c | 11 ++++++----- drivers/phy/phy-exynos-dp-video.c | 24 ++++-------------------- drivers/phy/phy-exynos-mipi-video.c | 11 ++++------- drivers/phy/phy-exynos4210-usb2.c | 1 - drivers/phy/phy-exynos4x12-usb2.c | 1 - drivers/phy/phy-exynos5-usbdrd.c | 2 +- drivers/phy/phy-exynos5250-usb2.c | 1 - drivers/phy/phy-hix5hd2-sata.c | 3 +++ drivers/phy/phy-miphy28lp.c | 13 +++++++------ drivers/phy/phy-miphy365x.c | 12 ++++++------ drivers/phy/phy-omap-control.c | 2 +- drivers/phy/phy-omap-usb2.c | 7 ++++--- drivers/phy/phy-rockchip-usb.c | 6 +++--- drivers/phy/phy-ti-pipe3.c | 12 ++++-------- drivers/phy/phy-twl4030-usb.c | 1 - drivers/phy/phy-xgene.c | 1 - 17 files changed, 45 insertions(+), 66 deletions(-)
Signature · 1 line
1.7.9.5
PATCH 1/20 phy: miphy28lp: Avoid calling of_get_child_count() multiple times
From: Axel Lin <[email protected]>
Currently, of_get_child_count() is called in each iteration of the for loop in miphy28lp_xlate(). This patch stores the return value of of_get_child_count() in miphy_dev->nphys and call of_get_child_count() once in miphy28lp_probe().
--- drivers/phy/phy-miphy28lp.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/phy/phy-miphy28lp.c +6 -5
@@ -228,6 +228,7 @@ struct miphy28lp_dev {struct regmap *regmap;struct mutex miphy_mutex;struct miphy28lp_phy **phys; +int nphys; };struct miphy_initval {
@@ -1116,7 +1117,7 @@ static struct phy *miphy28lp_xlate(struct device *dev,return ERR_PTR (-EINVAL ); } -for (index =0 ; index <of_get_child_count (dev->of_node); index++) +for (index =0 ; index < miphy_dev->nphys; index++)if (phynode == miphy_dev->phys[index]->phy->dev.of_node) { miphy_phy = miphy_dev->phys[index];break ;
@@ -1200,15 +1201,15 @@ static int miphy28lp_probe(struct platform_device *pdev)struct miphy28lp_dev *miphy_dev;struct phy_provider *provider;struct phy *phy; -int chancount, port =0 ; -int ret; +int ret, port =0 ; miphy_dev =devm_kzalloc (&pdev->dev,sizeof (*miphy_dev),GFP_KERNEL );if (!miphy_dev)return -ENOMEM ; - chancount =of_get_child_count (np); - miphy_dev->phys =devm_kzalloc (&pdev->dev,sizeof (phy) * chancount, + miphy_dev->nphys =of_get_child_count (np); + miphy_dev->phys =devm_kzalloc (&pdev->dev, +sizeof (phy) * miphy_dev->nphys,GFP_KERNEL );if (!miphy_dev->phys)return -ENOMEM ;
Signature · 1 line
1.7.9.5
PATCH 2/20 phy: miphy365x: Avoid calling of_get_child_count() multiple times
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Currently, of_get_child_count() is called in each iteration of the for loop in miphy365x_xlate(). This patch stores the return value of of_get_child_count() in miphy_dev->nphys and call of_get_child_count() once in miphy365x_probe().
--- drivers/phy/phy-miphy365x.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/phy/phy-miphy365x.c +6 -5
@@ -150,6 +150,7 @@ struct miphy365x_dev {struct regmap *regmap;struct mutex miphy_mutex;struct miphy365x_phy **phys; +int nphys; };/*
@@ -485,7 +486,7 @@ static struct phy *miphy365x_xlate(struct device *dev,return ERR_PTR (-EINVAL ); } -for (index =0 ; index <of_get_child_count (dev->of_node); index++) +for (index =0 ; index < miphy_dev->nphys; index++)if (phynode == miphy_dev->phys[index]->phy->dev.of_node) { miphy_phy = miphy_dev->phys[index];break ;
@@ -541,15 +542,15 @@ static int miphy365x_probe(struct platform_device *pdev)struct miphy365x_dev *miphy_dev;struct phy_provider *provider;struct phy *phy; -int chancount, port =0 ; -int ret; +int ret, port =0 ; miphy_dev =devm_kzalloc (&pdev->dev,sizeof (*miphy_dev),GFP_KERNEL );if (!miphy_dev)return -ENOMEM ; - chancount =of_get_child_count (np); - miphy_dev->phys =devm_kzalloc (&pdev->dev,sizeof (phy) * chancount, + miphy_dev->nphys =of_get_child_count (np); + miphy_dev->phys =devm_kzalloc (&pdev->dev, +sizeof (phy) * miphy_dev->nphys,GFP_KERNEL );if (!miphy_dev->phys)return -ENOMEM ;
Signature · 1 line
1.7.9.5
PATCH 3/20 phy: armada375-usb2: Set drvdata for phy and use it
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
At the context we have pointer to struct phy, it's useful to call phy_get_drvdata() to get the address of cluster_phy. This has slightly better readability than calling dev_get_drvdata(phy->dev.parent).
--- drivers/phy/phy-armada375-usb2.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/phy/phy-armada375-usb2.c +2 -1
@@ -37,7 +37,7 @@ static int armada375_usb_phy_init(struct phy *phy)struct armada375_cluster_phy *cluster_phy;u32 reg; - cluster_phy =dev_get_drvdata (phy->dev.parent); + cluster_phy =phy_get_drvdata (phy);if (!cluster_phy)return -ENODEV ;
@@ -131,6 +131,7 @@ static int armada375_usb_phy_probe(struct platform_device *pdev) cluster_phy->reg = usb_cluster_base;dev_set_drvdata (dev, cluster_phy); +phy_set_drvdata (phy, cluster_phy); phy_provider =devm_of_phy_provider_register (&pdev->dev, armada375_usb_phy_xlate);
Signature · 1 line
1.7.9.5
PATCH 4/20 phy: xgene: Remove duplicate code to set ctx->dev
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Set it once is enough and it's done after devm_kzalloc().
--- drivers/phy/phy-xgene.c | 1 - 1 file changed, 1 deletion(-)
drivers/phy/phy-xgene.c +0 -1
@@ -1704,7 +1704,6 @@ static int xgene_phy_probe(struct platform_device *pdev)for (i =0 ; i <MAX_LANE ; i++) ctx->sata_param.speed[i] =2 ;/* Default to Gen3 */ - ctx->dev = &pdev->dev;platform_set_drvdata (pdev, ctx); ctx->phy =devm_phy_create (ctx->dev,NULL , &xgene_phy_ops);
Signature · 1 line
1.7.9.5
PATCH 5/20 phy: miphy28lp: Add missing .owner field in miphy28lp_ops
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Add missing .owner field in miphy28lp_ops, which is used for refcounting.
--- drivers/phy/phy-miphy28lp.c | 1 + 1 file changed, 1 insertion(+)
drivers/phy/phy-miphy28lp.c +1 -0
@@ -1139,6 +1139,7 @@ static struct phy *miphy28lp_xlate(struct device *dev,static struct phy_ops miphy28lp_ops = { .init = miphy28lp_init, + .owner =THIS_MODULE , }; static int miphy28lp_probe_resets(struct device_node *node,
Signature · 1 line
1.7.9.5
PATCH 6/20 phy: exynos-mipi-video: Fixup the test for state->regmap
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
syscon_regmap_lookup_by_phandle() returns ERR_PTR on error. Thus don't use null test against state->regmap.
--- drivers/phy/phy-exynos-mipi-video.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/phy-exynos-mipi-video.c +1 -1
@@ -59,7 +59,7 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state,else reset =EXYNOS4_MIPI_PHY_SRESETN ; -if (state->regmap) { +if (!IS_ERR (state->regmap)) {mutex_lock (&state->mutex);regmap_read (state->regmap, offset, &val);if (on)
Signature · 1 line
1.7.9.5
PATCH 7/20 phy: exynos-mipi-video: Use spin_lock to protct state->regmap rmw operations
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
The state->regmap is initialized by devm_regmap_init_mmio().
So it's fine to use spin_lock rather than mutex to protct state->regmap rmw operations.
[[email protected]: Found an issue with the original patch w.r.t unbalanced
spin_lock call]
--- drivers/phy/phy-exynos-mipi-video.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-)
drivers/phy/phy-exynos-mipi-video.c +3 -6
@@ -43,7 +43,6 @@ struct exynos_mipi_video_phy { } phys[EXYNOS_MIPI_PHYS_NUM ];spinlock_t slock;void __iomem *regs; -struct mutex mutex;struct regmap *regmap; };
@@ -59,8 +58,9 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state,else reset =EXYNOS4_MIPI_PHY_SRESETN ; +spin_lock (&state->slock); +if (!IS_ERR (state->regmap)) { -mutex_lock (&state->mutex);regmap_read (state->regmap, offset, &val);if (on) val |= reset;
@@ -72,11 +72,9 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state,else if (!(val &EXYNOS4_MIPI_PHY_RESET_MASK )) val &= ~EXYNOS4_MIPI_PHY_ENABLE ;regmap_write (state->regmap, offset, val); -mutex_unlock (&state->mutex); }else { addr = state->regs +EXYNOS_MIPI_PHY_CONTROL (id /2 ); -spin_lock (&state->slock); val =readl (addr);if (on) val |= reset;
@@ -90,9 +88,9 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state, val &= ~EXYNOS4_MIPI_PHY_ENABLE ;writel (val, addr); -spin_unlock (&state->slock); } +spin_unlock (&state->slock);return 0 ; }
@@ -158,7 +156,6 @@ static int exynos_mipi_video_phy_probe(struct platform_device *pdev)dev_set_drvdata (dev, state);spin_lock_init (&state->slock); -mutex_init (&state->mutex);for (i =0 ; i <EXYNOS_MIPI_PHYS_NUM ; i++) { struct phy *phy = devm_phy_create(dev, NULL,
Signature · 1 line
1.7.9.5
PATCH 8/20 phy: exynos-dp-video: Kill exynos_dp_video_phy_pwr_isol function
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
If IS_ERR(state->regs) the .probe fails.
So IS_ERR(state->regs) test in exynos_dp_video_phy_pwr_isol() is not necessary. exynos_dp_video_phy_pwr_isol() simply does a regmap_update_bits() call now, just call regmap_update_bits() instead and return proper return value.
--- drivers/phy/phy-exynos-dp-video.c | 24 ++++-------------------- 1 file changed, 4 insertions(+), 20 deletions(-)
drivers/phy/phy-exynos-dp-video.c +4 -20
@@ -30,28 +30,13 @@ struct exynos_dp_video_phy {const struct exynos_dp_video_phy_drvdata *drvdata; }; -static void exynos_dp_video_phy_pwr_isol (struct exynos_dp_video_phy *state, -unsigned int on) -{ -unsigned int val; - -if (IS_ERR (state->regs)) -return ; - - val = on ?0 :EXYNOS5_PHY_ENABLE ; - -regmap_update_bits (state->regs, state->drvdata->phy_ctrl_offset, -EXYNOS5_PHY_ENABLE , val); -} -static int exynos_dp_video_phy_power_on (struct phy *phy) {struct exynos_dp_video_phy *state =phy_get_drvdata (phy);/* Disable power isolation on DP-PHY */ -exynos_dp_video_phy_pwr_isol (state,0 ); - -return 0 ; +return regmap_update_bits (state->regs, state->drvdata->phy_ctrl_offset, +EXYNOS5_PHY_ENABLE ,EXYNOS5_PHY_ENABLE ); }static int exynos_dp_video_phy_power_off (struct phy *phy)
@@ -59,9 +44,8 @@ static int exynos_dp_video_phy_power_off(struct phy *phy)struct exynos_dp_video_phy *state =phy_get_drvdata (phy);/* Enable power isolation on DP-PHY */ -exynos_dp_video_phy_pwr_isol (state,1 ); - -return 0 ; +return regmap_update_bits (state->regs, state->drvdata->phy_ctrl_offset, +EXYNOS5_PHY_ENABLE ,0 ); }static struct phy_ops exynos_dp_video_phy_ops = {
Signature · 1 line
1.7.9.5
PATCH 9/20 phy: hix5hd2-sata: Check return value of platform_get_resource
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
This prevent NULL pointer dereference if res is NULL.
--- drivers/phy/phy-hix5hd2-sata.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/phy/phy-hix5hd2-sata.c +3 -0
@@ -147,6 +147,9 @@ static int hix5hd2_sata_phy_probe(struct platform_device *pdev)return -ENOMEM ; res =platform_get_resource (pdev,IORESOURCE_MEM ,0 ); +if (!res) +return -EINVAL ; + priv->base =devm_ioremap (dev, res->start,resource_size (res));if (!priv->base)return -ENOMEM ;
Signature · 1 line
1.7.9.5
PATCH 10/20 phy: samsung-usb2: Remove NULL terminating entry from phys array
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Current code uses num_phys settings to tell the number of entries in phys. Thus remove the NULL terminating entry from phys array which is not necessary.
--- drivers/phy/phy-exynos4210-usb2.c | 1 - drivers/phy/phy-exynos4x12-usb2.c | 1 - drivers/phy/phy-exynos5250-usb2.c | 1 - 3 files changed, 3 deletions(-)
drivers/phy/phy-exynos4210-usb2.c +0 -1
@@ -250,7 +250,6 @@ static const struct samsung_usb2_common_phy exynos4210_phys[] = { .power_on = exynos4210_power_on, .power_off = exynos4210_power_off, }, - {}, }; const struct samsung_usb2_phy_config exynos4210_usb2_phy_config = {
drivers/phy/phy-exynos4x12-usb2.c +0 -1
@@ -361,7 +361,6 @@ static const struct samsung_usb2_common_phy exynos4x12_phys[] = { .power_on = exynos4x12_power_on, .power_off = exynos4x12_power_off, }, - {}, }; const struct samsung_usb2_phy_config exynos3250_usb2_phy_config = {
drivers/phy/phy-exynos5250-usb2.c +0 -1
@@ -391,7 +391,6 @@ static const struct samsung_usb2_common_phy exynos5250_phys[] = { .power_on = exynos5250_power_on, .power_off = exynos5250_power_off, }, - {}, }; const struct samsung_usb2_phy_config exynos5250_usb2_phy_config = {
Signature · 1 line
1.7.9.5
PATCH 11/20 phy: ti-pipe3: Simplify ti_pipe3_dpll_wait_lock implementation
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Code simplification. No functional change.
--- drivers/phy/phy-ti-pipe3.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-)
drivers/phy/phy-ti-pipe3.c +3 -7
@@ -165,15 +165,11 @@ static int ti_pipe3_dpll_wait_lock(struct ti_pipe3 *phy)cpu_relax (); val =ti_pipe3_readl (phy->pll_ctrl_base,PLL_STATUS );if (val &PLL_LOCK ) -break ; +return 0 ; }while (!time_after (jiffies, timeout)); -if (!(val &PLL_LOCK )) { -dev_err (phy->dev,"DPLL failed to lock\n" ); -return -EBUSY ; - } - -return 0 ; +dev_err (phy->dev,"DPLL failed to lock\n" ); +return -EBUSY ; }static int ti_pipe3_dpll_program (struct ti_pipe3 *phy)
Signature · 1 line
1.7.9.5
PATCH 12/20 phy: rockchip-usb: Fixup rockchip_usb_phy_power_on failure path
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
If rockchip_usb_phy_power() fails, we need to call clk_disable_unprepare() before return. This is to ensure we have balanced clk_enable/disable calls. Also remove unneeded ret checking in rockchip_usb_phy_power_off.
--- drivers/phy/phy-rockchip-usb.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/phy/phy-rockchip-usb.c +3 -3
@@ -61,8 +61,6 @@ static int rockchip_usb_phy_power_off(struct phy *_phy)return ret;clk_disable_unprepare (phy->clk); -if (ret) -return ret;return 0 ; }
@@ -78,8 +76,10 @@ static int rockchip_usb_phy_power_on(struct phy *_phy)/* Power up usb phy analog blocks by set siddq 0 */ ret =rockchip_usb_phy_power (phy,0 ); -if (ret) +if (ret) { +clk_disable_unprepare (phy->clk);return ret; + }return 0 ; }
Signature · 1 line
1.7.9.5
PATCH 13/20 phy: Find the right match in devm_phy_destroy()
In reply to Kishon Vijay Abraham IFrom: Thierry Reding <[email protected]>
devm_phy_create() stores the pointer to the new PHY at the address returned by devres_alloc(). The res parameter passed to devm_phy_match() is therefore the location where the pointer to the PHY is stored, hence it needs to be dereferenced before comparing to the match data in order to find the correct match.
--- drivers/phy/phy-core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/phy/phy-core.c +3 -1
@@ -52,7 +52,9 @@ static void devm_phy_consume(struct device *dev, void *res)static int devm_phy_match (struct device *dev,void *res,void *match_data) { -return res == match_data; +struct phy **phy = res; + +return *phy == match_data; }/**
Signature · 1 line
1.7.9.5
PATCH 14/20 phy: exynos5-usbdrd: Fix off-by-one valid value checking for args->args[0]
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Current code uses args->args[0] as array subscript of phy_drd->phys[]. So the valid value range for args->args[0] is 0 ... EXYNOS5_DRDPHYS_NUM - 1.
Reviewed by: Vivek Gautam <[email protected]>
--- drivers/phy/phy-exynos5-usbdrd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/phy-exynos5-usbdrd.c +1 -1
@@ -531,7 +531,7 @@ static struct phy *exynos5_usbdrd_phy_xlate(struct device *dev, {struct exynos5_usbdrd_phy *phy_drd =dev_get_drvdata (dev); -if (WARN_ON (args->args[0 ] >EXYNOS5_DRDPHYS_NUM )) +if (WARN_ON (args->args[0 ] >=EXYNOS5_DRDPHYS_NUM ))return ERR_PTR (-ENODEV );return phy_drd->phys[args->args[0 ]].phy;
Signature · 1 line
1.7.9.5
PATCH 15/20 phy: twl4030-usb: Remove redundant assignment for twl->linkstat
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
It's pointless to set twl->linkstat twice.
--- drivers/phy/phy-twl4030-usb.c | 1 - 1 file changed, 1 deletion(-)
drivers/phy/phy-twl4030-usb.c +0 -1
@@ -666,7 +666,6 @@ static int twl4030_usb_probe(struct platform_device *pdev) twl->dev = &pdev->dev; twl->irq =platform_get_irq (pdev,0 ); twl->vbus_supplied = false; - twl->linkstat = -EINVAL ; twl->linkstat =OMAP_MUSB_UNKNOWN ; twl->phy.dev = twl->dev;
Signature · 1 line
1.7.9.5
PATCH 16/20 phy: miphy365x: Convert to devm_kcalloc and fix wrong sizeof
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Prefer devm_kcalloc over devm_kzalloc with multiply. In additional, use sizeof(phy) is incorrect, fix it.
--- drivers/phy/phy-miphy365x.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/phy/phy-miphy365x.c +2 -3
@@ -549,9 +549,8 @@ static int miphy365x_probe(struct platform_device *pdev)return -ENOMEM ; miphy_dev->nphys =of_get_child_count (np); - miphy_dev->phys =devm_kzalloc (&pdev->dev, -sizeof (phy) * miphy_dev->nphys, -GFP_KERNEL ); + miphy_dev->phys =devm_kcalloc (&pdev->dev, miphy_dev->nphys, +sizeof (*miphy_dev->phys),GFP_KERNEL );if (!miphy_dev->phys)return -ENOMEM ;
Signature · 1 line
1.7.9.5
PATCH 17/20 phy: miphy28lp: Convert to devm_kcalloc and fix wrong sizof
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Prefer devm_kcalloc over devm_kzalloc with multiply. In additional, use sizeof(phy) is incorrect, fix it.
--- drivers/phy/phy-miphy28lp.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/phy/phy-miphy28lp.c +2 -3
@@ -1209,9 +1209,8 @@ static int miphy28lp_probe(struct platform_device *pdev)return -ENOMEM ; miphy_dev->nphys =of_get_child_count (np); - miphy_dev->phys =devm_kzalloc (&pdev->dev, -sizeof (phy) * miphy_dev->nphys, -GFP_KERNEL ); + miphy_dev->phys =devm_kcalloc (&pdev->dev, miphy_dev->nphys, +sizeof (*miphy_dev->phys),GFP_KERNEL );if (!miphy_dev->phys)return -ENOMEM ;
Signature · 1 line
1.7.9.5
PATCH 18/20 phy: core: Fixup return value of phy_exit when !pm_runtime_enabled
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
When phy_pm_runtime_get_sync() returns -ENOTSUPP, phy_exit() also returns -ENOTSUPP if !phy->ops->exit. Fix it.
Also move the code to override ret close to the code we got ret. I think it is less error prone this way.
--- drivers/phy/phy-core.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/phy/phy-core.c +3 -4
@@ -225,6 +225,7 @@ int phy_init(struct phy *phy) ret =phy_pm_runtime_get_sync (phy);if (ret <0 && ret != -ENOTSUPP )return ret; + ret =0 ;/* Override possible ret == -ENOTSUPP */ mutex_lock (&phy->mutex);if (phy->init_count ==0 && phy->ops->init) {
@@ -233,8 +234,6 @@ int phy_init(struct phy *phy)dev_err (&phy->dev,"phy init failed --> %d\n" , ret); goto out; } - }else { - ret =0 ;/* Override possible ret == -ENOTSUPP */ } ++phy->init_count;
@@ -255,6 +254,7 @@ int phy_exit(struct phy *phy) ret =phy_pm_runtime_get_sync (phy);if (ret <0 && ret != -ENOTSUPP )return ret; + ret =0 ;/* Override possible ret == -ENOTSUPP */ mutex_lock (&phy->mutex);if (phy->init_count ==1 && phy->ops->exit) {
@@ -289,6 +289,7 @@ int phy_power_on(struct phy *phy) ret =phy_pm_runtime_get_sync (phy);if (ret <0 && ret != -ENOTSUPP )return ret; + ret =0 ;/* Override possible ret == -ENOTSUPP */ mutex_lock (&phy->mutex);if (phy->power_count ==0 && phy->ops->power_on) {
@@ -297,8 +298,6 @@ int phy_power_on(struct phy *phy)dev_err (&phy->dev,"phy poweron failed --> %d\n" , ret); goto out; } - }else { - ret =0 ;/* Override possible ret == -ENOTSUPP */ } ++phy->power_count;mutex_unlock (&phy->mutex);
Signature · 1 line
1.7.9.5
PATCH 19/20 phy: ti/omap: Fix modalias
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Remove extra space in MODULE_ALIAS.
--- drivers/phy/phy-omap-control.c | 2 +- drivers/phy/phy-omap-usb2.c | 2 +- drivers/phy/phy-ti-pipe3.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)
drivers/phy/phy-omap-control.c +1 -1
@@ -360,7 +360,7 @@ static void __exit omap_control_phy_exit(void) }module_exit (omap_control_phy_exit); -MODULE_ALIAS ("platform: omap_control_phy" ); +MODULE_ALIAS ("platform:omap_control_phy" );MODULE_AUTHOR ("Texas Instruments Inc." );MODULE_DESCRIPTION ("OMAP Control Module PHY Driver" );MODULE_LICENSE ("GPL v2" );
drivers/phy/phy-omap-usb2.c +1 -1
@@ -383,7 +383,7 @@ static struct platform_driver omap_usb2_driver = {module_platform_driver (omap_usb2_driver); -MODULE_ALIAS ("platform: omap_usb2" ); +MODULE_ALIAS ("platform:omap_usb2" );MODULE_AUTHOR ("Texas Instruments Inc." );MODULE_DESCRIPTION ("OMAP USB2 phy driver" );MODULE_LICENSE ("GPL v2" );
drivers/phy/phy-ti-pipe3.c +1 -1
@@ -604,7 +604,7 @@ static struct platform_driver ti_pipe3_driver = {module_platform_driver (ti_pipe3_driver); -MODULE_ALIAS ("platform: ti_pipe3" ); +MODULE_ALIAS ("platform:ti_pipe3" );MODULE_AUTHOR ("Texas Instruments Inc." );MODULE_DESCRIPTION ("TI PIPE3 phy driver" );MODULE_LICENSE ("GPL v2" );
Signature · 1 line
1.7.9.5
PATCH 20/20 phy: omap-usb2: Fix missing clk_prepare call when using old dt name
In reply to Kishon Vijay Abraham IFrom: Axel Lin <[email protected]>
Current code does not call clk_prepare(phy->optclk) when using the old usb_otg_ss_refclk960m name. Fix it.
--- drivers/phy/phy-omap-usb2.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/phy/phy-omap-usb2.c +3 -2
@@ -296,10 +296,11 @@ static int omap_usb2_probe(struct platform_device *pdev)dev_warn (&pdev->dev,"found usb_otg_ss_refclk960m, please fix DTS\n" ); } - }else { -clk_prepare (phy->optclk); } +if (!IS_ERR (phy->optclk)) +clk_prepare (phy->optclk); +usb_add_phy_dev (&phy->phy);return 0 ;
Signature · 1 line
1.7.9.5
Quoted 17 lines · Kishon Vijay Abraham I
Hi Greg,
Please find the pull request for 4.0 -rc cycle. We should thank Axel Lin for reviewing all the PHY drivers and sending fixes. This also includes a fix from Thierry Reding in phy core.
Let me know if I have to change something.
Thanks
KishonThe following changes since commit
c517d838eb7d07bbe9507871fab3931deccff539:Linux 4.0-rc1 (2015-02-22 18:21:14 -0800)
are available in the git repository at:
Quoted 9 lines
git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux-phy.gittags/for-4.0-rcfor you to fetch changes up to
b1ff3231b2d4197ef5024f9c57ffc6cfa562590c:phy: omap-usb2: Fix missing
clk_preparecall when using old dt name (2015-03-13 17:14:39 +0530)----------------------------------------------------------------
contains fixes all overdrivers/phy
That's a horrible pull request summary. Please at least give me something "real" to go on here. As it is, would you accept such a changelog entry?
I'll just take the individual patches, to keep this out of the commit log...
thanks,
greg k-h
Hi Greg,
Quoted 28 lines · Greg KH
Quoted 17 lines · Kishon Vijay Abraham I
Hi Greg,
Please find the pull request for 4.0 -rc cycle. We should thank Axel Lin for reviewing all the PHY drivers and sending fixes. This also includes a fix from Thierry Reding in phy core.
Let me know if I have to change something.
Thanks
KishonThe following changes since commit
c517d838eb7d07bbe9507871fab3931deccff539:Linux 4.0-rc1 (2015-02-22 18:21:14 -0800)are available in the git repository at:
Quoted 9 lines
git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux-phy.gittags/for-4.0-rcfor you to fetch changes up to
b1ff3231b2d4197ef5024f9c57ffc6cfa562590c:phy: omap-usb2: Fix missing
clk_preparecall when using old dt name (2015-03-13 17:14:39 +0530)----------------------------------------------------------------
contains fixes all overdrivers/phy
Quoted 6 lines
That's a horrible pull request summary. Please at least give me something "real" to go on here. As it is, would you accept such a changelog entry?
I'll just take the individual patches, to keep this out of the commit log...
sorry, I'll send an updated pull request if that is okay.
Thanks
Kishon
Hi Greg,
Please find the updated pull request for 4.0 -rc. This mostly consists of fixes done by Axel Lin all over the PHY drivers. It also includes a couple of fixes in phy core.
Let me know if I have to change something.
Thanks
Kishon
The following changes since commit c517d838eb7d07bbe9507871fab3931deccff539:
Linux 4.0-rc1 (2015-02-22 18:21:14 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux-phy.git
tags/for-4.0-rc
for you to fetch changes up to b1ff3231b2d4197ef5024f9c57ffc6cfa562590c:
phy: omap-usb2: Fix missing clk_prepare call when using old dt name
(2015-03-13 17:14:39 +0530)
----------------------------------------------------------------
contains fixes all over drivers/phy and includes the following *) Using phy_get_drvdata instead of dev_get_drvdata in armada375-usb2.c *) Fixes w.r.t checking return values in regmap APIs, protecting regmap ops
with spin lock and directly using regmap_update_bits instead of having a separate function to do the same in the various PHYs used in exynos.
*) check return value in platform_get_resource of hix5hd2-sata PHY *) Removed NULL terminating entry from phys array and fix off-by-one
valid value checking for args->args[0] in of_xlate of exynos USB PHY.
*) Fixup rockchip_usb_phy_power_on failure path
*) Fix devm_phy_match to find the correct match in phy core and also fix to
return the correct value from PHY APIs if PM runtime is not enabled.
*) Removed redundant code in twl4030 and xgene
*) Fixed sizeof during memory allocation in miphy PHYs
*) simplify ti_pipe3_dpll_wait_lock implementation, fix missing clk_prepare
when using old dt name and nit pick in MOUDLE_ALIAS in TI PHYs.
----------------------------------------------------------------
Axel Lin (19):
phy: miphy28lp: Avoid calling of_get_child_count() multiple times
phy: miphy365x: Avoid calling of_get_child_count() multiple times
phy: armada375-usb2: Set drvdata for phy and use it
phy: xgene: Remove duplicate code to set ctx->dev
phy: miphy28lp: Add missing .owner field in miphy28lp_ops
phy: exynos-mipi-video: Fixup the test for state->regmap
phy: exynos-mipi-video: Use spin_lock to protct state->regmap rmw operations
phy: exynos-dp-video: Kill exynos_dp_video_phy_pwr_isol function
phy: hix5hd2-sata: Check return value of platform_get_resource
phy: samsung-usb2: Remove NULL terminating entry from phys array
phy: ti-pipe3: Simplify ti_pipe3_dpll_wait_lock implementation
phy: rockchip-usb: Fixup rockchip_usb_phy_power_on failure path
phy: exynos5-usbdrd: Fix off-by-one valid value checking for args->args[0]
phy: twl4030-usb: Remove redundant assignment for twl->linkstat
phy: miphy365x: Convert to devm_kcalloc and fix wrong sizeof
phy: miphy28lp: Convert to devm_kcalloc and fix wrong sizof
phy: core: Fixup return value of phy_exit when !pm_runtime_enabled
phy: ti/omap: Fix modalias
phy: omap-usb2: Fix missing clk_prepare call when using old dt name
Thierry Reding (1):
phy: Find the right match in devm_phy_destroy()
drivers/phy/phy-armada375-usb2.c | 3 ++- drivers/phy/phy-core.c | 11 ++++++----- drivers/phy/phy-exynos-dp-video.c | 24 ++++-------------------- drivers/phy/phy-exynos-mipi-video.c | 11 ++++------- drivers/phy/phy-exynos4210-usb2.c | 1 - drivers/phy/phy-exynos4x12-usb2.c | 1 - drivers/phy/phy-exynos5-usbdrd.c | 2 +- drivers/phy/phy-exynos5250-usb2.c | 1 - drivers/phy/phy-hix5hd2-sata.c | 3 +++ drivers/phy/phy-miphy28lp.c | 13 +++++++------ drivers/phy/phy-miphy365x.c | 12 ++++++------ drivers/phy/phy-omap-control.c | 2 +- drivers/phy/phy-omap-usb2.c | 7 ++++--- drivers/phy/phy-rockchip-usb.c | 6 +++--- drivers/phy/phy-ti-pipe3.c | 12 ++++-------- drivers/phy/phy-twl4030-usb.c | 1 - drivers/phy/phy-xgene.c | 1 - 17 files changed, 45 insertions(+), 66 deletions(-)
Quoted 11 lines · Kishon Vijay Abraham I
Hi Greg,
Please find the updated pull request for 4.0 -rc. This mostly consists of fixes done by Axel Lin all over the PHY drivers. It also includes a couple of fixes in phy core.
Let me know if I have to change something.
Thanks
Kishon
Quoted 8 lines
The following changes since commit
c517d838eb7d07bbe9507871fab3931deccff539:Linux 4.0-rc1 (2015-02-22 18:21:14 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux-phy.git
tags/for-4.0-rc
All pulled and pushed out, thanks.
greg k-h