net: macb: Not all the GEM are gigabit capable

PATCH New Gregory CLEMENT 4 participants 5 messages Dec 16, 2015 – Jan 27, 2016 Network drivers

Share & tools

Gregory CLEMENT

author

2015-12-16 13:50 UTC #0 lore raw

PATCH v3 1/1 net: macb: Not all the GEM are gigabit capable

During the initialization of PHY the gigabit bit capable is set if the controller is a GEM. However, for sama5d2 and sama5d4, the GEM is configured to support only 10/100.

Improperly setting the GBE capability leads to an unresponsive MAC controller. This patch fixes this behavior allowing using the gmac with these SoCs.

Signed-off-by Gregory CLEMENT <[email protected]>

---
Hi,

in this v3, I fixed the sama5d2 typo and the missing brace.

I also took care of the cpu_is_* symbols which may not be defined depending of the choice of the SoC.

Gregory

drivers/net/macb.c | 21 +++++++++++++++++++--

 1 file changed, 19 insertions(+), 2 deletions(-)

drivers/net/macb.c +19 -2

index a5c1880..e801313 100644
@@ -109,6 +109,23 @@ static int macb_is_gem(struct macb_device *macb)
 	return MACB_BFEXT(IDNUM, macb_readl(macb, MID)) == 0x2;
 }
 
+#ifndef cpu_is_sama5d2
+#define cpu_is_sama5d2() 0
+#endif
+
+#ifndef cpu_is_sama5d4
+#define cpu_is_sama5d4() 0
+#endif
+
+static int gem_is_gigabit_capable(struct macb_device *macb)
+{
+	/*
+	 * The GEM controllers embeded in SAMA5D2 and SAMA5D4 are
+	 * configured to support only 10/100.
+	 */
+	return macb_is_gem(macb) && !cpu_is_sama5d2() && !cpu_is_sama5d4();
+}
+
 static void macb_mdio_write(struct macb_device *macb, u8 reg, u16 value)
 {
 	unsigned long netctl;

@@ -480,8 +497,8 @@ static int macb_phy_init(struct macb_device *macb)
 		return 0;
 	}
 
-	/* First check for GMAC */
-	if (macb_is_gem(macb)) {
+	/* First check for GMAC and that it is GiB capable */
+	if (gem_is_gigabit_capable(macb)){
 		lpa = macb_mdio_read(macb, MII_STAT1000);
 
 		if (lpa & (LPA_1000FULL | LPA_1000HALF)) {

Signature · 1 line

2.5.0

Nicolas Ferre

2015-12-16 14:04 UTC #1 lore raw

PATCH v3 1/1 net: macb: Not all the GEM are gigabit capable

Le 16/12/2015 14:50, Gregory CLEMENT a ?crit :

Quoted 10 lines

During the initialization of PHY the gigabit bit capable is set if the controller is a GEM. However, for sama5d2 and sama5d4, the GEM is configured to support only 10/100.
Improperly setting the GBE capability leads to an unresponsive MAC controller. This patch fixes this behavior allowing using the gmac with these SoCs.
Suggested-by Nicolas Ferre <[email protected]>
Signed-off-by Gregory CLEMENT <[email protected]>

Acked-by Nicolas Ferre <[email protected]>

Thanks,

Quoted 53 lines

---
Hi,

in this v3, I fixed the sama5d2 typo and the missing brace.

I also took care of the cpu_is_* symbols which may not be defined depending of the choice of the SoC.

Gregory

drivers/net/macb.c | 21 +++++++++++++++++++--

 1 file changed, 19 insertions(+), 2 deletions(-)

drivers/net/macb.c +19 -2

index a5c1880..e801313 100644
@@ -109,6 +109,23 @@ static int macb_is_gem(struct macb_device *macb)
 	return MACB_BFEXT(IDNUM, macb_readl(macb, MID)) == 0x2;
 }
 
+#ifndef cpu_is_sama5d2
+#define cpu_is_sama5d2() 0
+#endif
+
+#ifndef cpu_is_sama5d4
+#define cpu_is_sama5d4() 0
+#endif
+
+static int gem_is_gigabit_capable(struct macb_device *macb)
+{
+	/*
+	 * The GEM controllers embeded in SAMA5D2 and SAMA5D4 are
+	 * configured to support only 10/100.
+	 */
+	return macb_is_gem(macb) && !cpu_is_sama5d2() && !cpu_is_sama5d4();
+}
+
 static void macb_mdio_write(struct macb_device *macb, u8 reg, u16 value)
 {
 	unsigned long netctl;

@@ -480,8 +497,8 @@ static int macb_phy_init(struct macb_device *macb)
 		return 0;
 	}
 
-	/* First check for GMAC */
-	if (macb_is_gem(macb)) {
+	/* First check for GMAC and that it is GiB capable */
+	if (gem_is_gigabit_capable(macb)){
 		lpa = macb_mdio_read(macb, MII_STAT1000);
 
 		if (lpa & (LPA_1000FULL | LPA_1000HALF)) {

Signature · 1 line

Nicolas Ferre

Joe Hershberger

2015-12-16 21:06 UTC #2 lore raw

PATCH v3 1/1 net: macb: Not all the GEM are gigabit capable

In reply to Gregory CLEMENT

Quoted 10 lines · Gregory CLEMENT

During the initialization of PHY the gigabit bit capable is set if the controller is a GEM. However, for sama5d2 and sama5d4, the GEM is configured to support only 10/100.
Improperly setting the GBE capability leads to an unresponsive MAC controller. This patch fixes this behavior allowing using the gmac with these SoCs.
Suggested-by Nicolas Ferre <[email protected]>
Signed-off-by Gregory CLEMENT <[email protected]>

It's a shame we are forced to have this PID hack, but I guess this will do until we discover some other sustainable way.

Acked-by Joe Hershberger <[email protected]>

Andreas Bießmann

2016-01-24 21:53 UTC #3 lore raw

PATCH v3 1/1 net: macb: Not all the GEM are gigabit capable

In reply to Gregory CLEMENT

Quoted 10 lines · Gregory CLEMENT

During the initialization of PHY the gigabit bit capable is set if the controller is a GEM. However, for sama5d2 and sama5d4, the GEM is configured to support only 10/100.
Improperly setting the GBE capability leads to an unresponsive MAC controller. This patch fixes this behavior allowing using the gmac with these SoCs.
Suggested-by Nicolas Ferre <[email protected]>
Signed-off-by Gregory CLEMENT <[email protected]>

Reviewed-by Andreas Bie?mann <[email protected]>

Quoted 53 lines

---
Hi,

in this v3, I fixed the sama5d2 typo and the missing brace.

I also took care of the cpu_is_* symbols which may not be defined depending of the choice of the SoC.

Gregory

drivers/net/macb.c | 21 +++++++++++++++++++--

 1 file changed, 19 insertions(+), 2 deletions(-)

drivers/net/macb.c +19 -2

index a5c1880..e801313 100644
@@ -109,6 +109,23 @@ static int macb_is_gem(struct macb_device *macb)
 	return MACB_BFEXT(IDNUM, macb_readl(macb, MID)) == 0x2;
 }
 
+#ifndef cpu_is_sama5d2
+#define cpu_is_sama5d2() 0
+#endif
+
+#ifndef cpu_is_sama5d4
+#define cpu_is_sama5d4() 0
+#endif
+
+static int gem_is_gigabit_capable(struct macb_device *macb)
+{
+	/*
+	 * The GEM controllers embeded in SAMA5D2 and SAMA5D4 are
+	 * configured to support only 10/100.
+	 */
+	return macb_is_gem(macb) && !cpu_is_sama5d2() && !cpu_is_sama5d4();
+}
+
 static void macb_mdio_write(struct macb_device *macb, u8 reg, u16 value)
 {
 	unsigned long netctl;

@@ -480,8 +497,8 @@ static int macb_phy_init(struct macb_device *macb)
 		return 0;
 	}
 
-	/* First check for GMAC */
-	if (macb_is_gem(macb)) {
+	/* First check for GMAC and that it is GiB capable */
+	if (gem_is_gigabit_capable(macb)){
 		lpa = macb_mdio_read(macb, MII_STAT1000);
 
 		if (lpa & (LPA_1000FULL | LPA_1000HALF)) {

Andreas Bießmann

2016-01-27 13:00 UTC #4 lore raw

PATCH v3 1/1 net: macb: Not all the GEM are gigabit capable

In reply to Gregory CLEMENT

Dear Gregory CLEMENT,

Quoted 20 lines · Gregory CLEMENT

During the initialization of PHY the gigabit bit capable is set if the controller is a GEM. However, for sama5d2 and sama5d4, the GEM is configured to support only 10/100.
Improperly setting the GBE capability leads to an unresponsive MAC controller. This patch fixes this behavior allowing using the gmac with these SoCs.
Suggested-by Nicolas Ferre <[email protected]>
Signed-off-by Gregory CLEMENT <[email protected]>
Acked-by Nicolas Ferre <[email protected]>
Acked-by Joe Hershberger <[email protected]>
Reviewed-by Andreas Bie?mann <[email protected]>
[fixed minor checkpatch warning]
Signed-off-by Andreas Bie?mann <[email protected]>
---
Hi,
in this v3, I fixed the sama5d2 typo and the missing brace.

Quoted 7 lines

I also took care of the cpu_is_* symbols which may not be defined depending of the choice of the SoC.
Gregory
drivers/net/macb.c | 21 +++++++++++++++++++--
1 file changed, 19 insertions(+), 2 deletions(-)

applied to u-boot-atmel/master, thanks!

Best regards,
Andreas Bie?mann