* [PATCH v5 14/17] dt-bindings: slave-device: add current-speed property
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
This adds a new DT property to define the current baud rate of the
slave device.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
Documentation/devicetree/bindings/serial/slave-device.txt | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/Documentation/devicetree/bindings/serial/slave-device.txt b/Documentation/devicetree/bindings/serial/slave-device.txt
index f660379..40110e0 100644
--- a/Documentation/devicetree/bindings/serial/slave-device.txt
+++ b/Documentation/devicetree/bindings/serial/slave-device.txt
@@ -21,6 +21,15 @@ Optional Properties:
can support. For example, a particular board has some signal
quality issue or the host processor can't support higher
baud rates.
+- current-speed : The current baud rate the device operates at. This should
+ only be present in case a driver has no chance to know
+ the baud rate of the slave device.
+ Examples:
+ * device supports auto-baud
+ * the rate is setup by a bootloader and there is no
+ way to reset the device
+ * device baud rate is configured by its firmware but
+ there is no way to request the actual settings
Example:
--
2.1.4
^ permalink raw reply related
* [PATCH v5 13/17] dt-bindings: qca7000: rename binding
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
Before we can merge the QCA7000 UART binding the document needs to be
renamed.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
.../devicetree/bindings/net/{qca-qca7000-spi.txt => qca-qca7000.txt} | 0
1 file changed, 0 insertions(+), 0 deletions(-)
rename Documentation/devicetree/bindings/net/{qca-qca7000-spi.txt => qca-qca7000.txt} (100%)
diff --git a/Documentation/devicetree/bindings/net/qca-qca7000-spi.txt b/Documentation/devicetree/bindings/net/qca-qca7000.txt
similarity index 100%
rename from Documentation/devicetree/bindings/net/qca-qca7000-spi.txt
rename to Documentation/devicetree/bindings/net/qca-qca7000.txt
--
2.1.4
^ permalink raw reply
* [PATCH v5 12/17] dt-bindings: qca7000-spi: Rework binding
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
In preparation for the QCA7000 UART binding rework the binding document.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
.../devicetree/bindings/net/qca-qca7000-spi.txt | 49 +++++++++++++---------
1 file changed, 29 insertions(+), 20 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/qca-qca7000-spi.txt b/Documentation/devicetree/bindings/net/qca-qca7000-spi.txt
index c74989c..a37f656 100644
--- a/Documentation/devicetree/bindings/net/qca-qca7000-spi.txt
+++ b/Documentation/devicetree/bindings/net/qca-qca7000-spi.txt
@@ -1,29 +1,38 @@
-* Qualcomm QCA7000 (Ethernet over SPI protocol)
+* Qualcomm QCA7000
-Note: The QCA7000 is useable as a SPI device. In this case it must be defined
-as a child of a SPI master in the device tree.
+The QCA7000 is a serial-to-powerline bridge with a host interface which could
+be configured either as SPI or UART slave. This configuration is done by
+the QCA7000 firmware.
+
+(a) Ethernet over SPI
+
+In order to use the QCA7000 as SPI device it must be defined as a child of a
+SPI master in the device tree.
Required properties:
-- compatible : Should be "qca,qca7000"
-- reg : Should specify the SPI chip select
-- interrupts : The first cell should specify the index of the source interrupt
- and the second cell should specify the trigger type as rising edge
-- spi-cpha : Must be set
-- spi-cpol: Must be set
+- compatible : Should be "qca,qca7000"
+- reg : Should specify the SPI chip select
+- interrupts : The first cell should specify the index of the source
+ interrupt and the second cell should specify the trigger
+ type as rising edge
+- spi-cpha : Must be set
+- spi-cpol : Must be set
Optional properties:
-- interrupt-parent : Specify the pHandle of the source interrupt
+- interrupt-parent : Specify the pHandle of the source interrupt
- spi-max-frequency : Maximum frequency of the SPI bus the chip can operate at.
- Numbers smaller than 1000000 or greater than 16000000 are invalid. Missing
- the property will set the SPI frequency to 8000000 Hertz.
-- local-mac-address: 6 bytes, MAC address
-- qca,legacy-mode : Set the SPI data transfer of the QCA7000 to legacy mode.
- In this mode the SPI master must toggle the chip select between each data
- word. In burst mode these gaps aren't necessary, which is faster.
- This setting depends on how the QCA7000 is setup via GPIO pin strapping.
- If the property is missing the driver defaults to burst mode.
-
-Example:
+ Numbers smaller than 1000000 or greater than 16000000
+ are invalid. Missing the property will set the SPI
+ frequency to 8000000 Hertz.
+- local-mac-address : see ./ethernet.txt
+- qca,legacy-mode : Set the SPI data transfer of the QCA7000 to legacy mode.
+ In this mode the SPI master must toggle the chip select
+ between each data word. In burst mode these gaps aren't
+ necessary, which is faster. This setting depends on how
+ the QCA7000 is setup via GPIO pin strapping. If the
+ property is missing the driver defaults to burst mode.
+
+SPI Example:
/* Freescale i.MX28 SPI master*/
ssp2: spi@80014000 {
--
2.1.4
^ permalink raw reply related
* [PATCH v5 11/17] net: qualcomm: make qca_7k_common a separate kernel module
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
In order to share common functions between QCA7000 SPI and UART protocol
driver the qca_7k_common needs to be a separate kernel module.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/Kconfig | 8 +++++++-
drivers/net/ethernet/qualcomm/Makefile | 5 +++--
drivers/net/ethernet/qualcomm/qca_7k_common.c | 10 ++++++++++
3 files changed, 20 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/Kconfig b/drivers/net/ethernet/qualcomm/Kconfig
index d7720bf..b4c369d 100644
--- a/drivers/net/ethernet/qualcomm/Kconfig
+++ b/drivers/net/ethernet/qualcomm/Kconfig
@@ -16,7 +16,13 @@ config NET_VENDOR_QUALCOMM
if NET_VENDOR_QUALCOMM
config QCA7000
- tristate "Qualcomm Atheros QCA7000 support"
+ tristate
+ help
+ This enables support for the Qualcomm Atheros QCA7000.
+
+config QCA7000_SPI
+ tristate "Qualcomm Atheros QCA7000 SPI support"
+ select QCA7000
depends on SPI_MASTER && OF
---help---
This SPI protocol driver supports the Qualcomm Atheros QCA7000.
diff --git a/drivers/net/ethernet/qualcomm/Makefile b/drivers/net/ethernet/qualcomm/Makefile
index 5e17bf1..65556ca 100644
--- a/drivers/net/ethernet/qualcomm/Makefile
+++ b/drivers/net/ethernet/qualcomm/Makefile
@@ -2,7 +2,8 @@
# Makefile for the Qualcomm network device drivers.
#
-obj-$(CONFIG_QCA7000) += qcaspi.o
-qcaspi-objs := qca_spi.o qca_7k_common.o qca_7k.o qca_debug.o
+obj-$(CONFIG_QCA7000) += qca_7k_common.o
+obj-$(CONFIG_QCA7000_SPI) += qcaspi.o
+qcaspi-objs := qca_7k.o qca_debug.o qca_spi.o
obj-y += emac/
diff --git a/drivers/net/ethernet/qualcomm/qca_7k_common.c b/drivers/net/ethernet/qualcomm/qca_7k_common.c
index 0d3daa9..6b511f0 100644
--- a/drivers/net/ethernet/qualcomm/qca_7k_common.c
+++ b/drivers/net/ethernet/qualcomm/qca_7k_common.c
@@ -21,7 +21,9 @@
* by an atheros frame while transmitted over a serial channel;
*/
+#include <linux/init.h>
#include <linux/kernel.h>
+#include <linux/module.h>
#include "qca_7k_common.h"
@@ -46,6 +48,7 @@ qcafrm_create_header(u8 *buf, u16 length)
return QCAFRM_HEADER_LEN;
}
+EXPORT_SYMBOL_GPL(qcafrm_create_header);
u16
qcafrm_create_footer(u8 *buf)
@@ -57,6 +60,7 @@ qcafrm_create_footer(u8 *buf)
buf[1] = 0x55;
return QCAFRM_FOOTER_LEN;
}
+EXPORT_SYMBOL_GPL(qcafrm_create_footer);
/* Gather received bytes and try to extract a full ethernet frame by
* following a simple state machine.
@@ -154,3 +158,9 @@ qcafrm_fsm_decode(struct qcafrm_handle *handle, u8 *buf, u16 buf_len, u8 recv_by
return ret;
}
+EXPORT_SYMBOL_GPL(qcafrm_fsm_decode);
+
+MODULE_DESCRIPTION("Qualcomm Atheros QCA7000 common");
+MODULE_AUTHOR("Qualcomm Atheros Communications");
+MODULE_AUTHOR("Stefan Wahren <stefan.wahren@i2se.com>");
+MODULE_LICENSE("Dual BSD/GPL");
--
2.1.4
^ permalink raw reply related
* [PATCH v5 09/17] net: qualcomm: rename qca_framing.c to qca_7k_common.c
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
As preparation for the upcoming UART driver we need a module
which contains common functions for both interfaces. The module
qca_framing is a good candidate but renaming to qca_7k_common would
make it clear.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/Makefile | 2 +-
drivers/net/ethernet/qualcomm/{qca_framing.c => qca_7k_common.c} | 2 +-
drivers/net/ethernet/qualcomm/{qca_framing.h => qca_7k_common.h} | 0
drivers/net/ethernet/qualcomm/qca_spi.c | 2 +-
drivers/net/ethernet/qualcomm/qca_spi.h | 2 +-
5 files changed, 4 insertions(+), 4 deletions(-)
rename drivers/net/ethernet/qualcomm/{qca_framing.c => qca_7k_common.c} (99%)
rename drivers/net/ethernet/qualcomm/{qca_framing.h => qca_7k_common.h} (100%)
diff --git a/drivers/net/ethernet/qualcomm/Makefile b/drivers/net/ethernet/qualcomm/Makefile
index aacb0a5..5e17bf1 100644
--- a/drivers/net/ethernet/qualcomm/Makefile
+++ b/drivers/net/ethernet/qualcomm/Makefile
@@ -3,6 +3,6 @@
#
obj-$(CONFIG_QCA7000) += qcaspi.o
-qcaspi-objs := qca_spi.o qca_framing.o qca_7k.o qca_debug.o
+qcaspi-objs := qca_spi.o qca_7k_common.o qca_7k.o qca_debug.o
obj-y += emac/
diff --git a/drivers/net/ethernet/qualcomm/qca_framing.c b/drivers/net/ethernet/qualcomm/qca_7k_common.c
similarity index 99%
rename from drivers/net/ethernet/qualcomm/qca_framing.c
rename to drivers/net/ethernet/qualcomm/qca_7k_common.c
index 2341f2b..6d17fbd 100644
--- a/drivers/net/ethernet/qualcomm/qca_framing.c
+++ b/drivers/net/ethernet/qualcomm/qca_7k_common.c
@@ -23,7 +23,7 @@
#include <linux/kernel.h>
-#include "qca_framing.h"
+#include "qca_7k_common.h"
u16
qcafrm_create_header(u8 *buf, u16 length)
diff --git a/drivers/net/ethernet/qualcomm/qca_framing.h b/drivers/net/ethernet/qualcomm/qca_7k_common.h
similarity index 100%
rename from drivers/net/ethernet/qualcomm/qca_framing.h
rename to drivers/net/ethernet/qualcomm/qca_7k_common.h
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
index deec70f..2bc3ba4 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.c
+++ b/drivers/net/ethernet/qualcomm/qca_spi.c
@@ -43,8 +43,8 @@
#include <linux/types.h>
#include "qca_7k.h"
+#include "qca_7k_common.h"
#include "qca_debug.h"
-#include "qca_framing.h"
#include "qca_spi.h"
#define MAX_DMA_BURST_LEN 5000
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.h b/drivers/net/ethernet/qualcomm/qca_spi.h
index 064853d..fc4beb1 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.h
+++ b/drivers/net/ethernet/qualcomm/qca_spi.h
@@ -32,7 +32,7 @@
#include <linux/spi/spi.h>
#include <linux/types.h>
-#include "qca_framing.h"
+#include "qca_7k_common.h"
#define QCASPI_DRV_VERSION "0.2.7-i"
#define QCASPI_DRV_NAME "qcaspi"
--
2.1.4
^ permalink raw reply related
* [PATCH v5 08/17] net: qca_spi: Clarify MODULE_DESCRIPTION
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
Since this driver is specific to the QCA7000, we should make the module
description more precisely.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_spi.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
index c727357..deec70f 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.c
+++ b/drivers/net/ethernet/qualcomm/qca_spi.c
@@ -997,7 +997,7 @@ static struct spi_driver qca_spi_driver = {
};
module_spi_driver(qca_spi_driver);
-MODULE_DESCRIPTION("Qualcomm Atheros SPI Driver");
+MODULE_DESCRIPTION("Qualcomm Atheros QCA7000 SPI Driver");
MODULE_AUTHOR("Qualcomm Atheros Communications");
MODULE_AUTHOR("Stefan Wahren <stefan.wahren@i2se.com>");
MODULE_LICENSE("Dual BSD/GPL");
--
2.1.4
^ permalink raw reply related
* [PATCH v5 07/17] net: qualcomm: move qcaspi_tx_cmd to qca_spi.c
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
The function qcaspi_tx_cmd() is only called from qca_spi.c. So we better
move it there.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_7k.c | 24 ------------------------
drivers/net/ethernet/qualcomm/qca_7k.h | 1 -
drivers/net/ethernet/qualcomm/qca_spi.c | 24 ++++++++++++++++++++++++
3 files changed, 24 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_7k.c b/drivers/net/ethernet/qualcomm/qca_7k.c
index 557d53c..aa90a1d 100644
--- a/drivers/net/ethernet/qualcomm/qca_7k.c
+++ b/drivers/net/ethernet/qualcomm/qca_7k.c
@@ -119,27 +119,3 @@ qcaspi_write_register(struct qcaspi *qca, u16 reg, u16 value)
return ret;
}
-
-int
-qcaspi_tx_cmd(struct qcaspi *qca, u16 cmd)
-{
- __be16 tx_data;
- struct spi_message *msg = &qca->spi_msg1;
- struct spi_transfer *transfer = &qca->spi_xfer1;
- int ret;
-
- tx_data = cpu_to_be16(cmd);
- transfer->len = sizeof(tx_data);
- transfer->tx_buf = &tx_data;
- transfer->rx_buf = NULL;
-
- ret = spi_sync(qca->spi_dev, msg);
-
- if (!ret)
- ret = msg->status;
-
- if (ret)
- qcaspi_spi_error(qca);
-
- return ret;
-}
diff --git a/drivers/net/ethernet/qualcomm/qca_7k.h b/drivers/net/ethernet/qualcomm/qca_7k.h
index 4047f0a..27124c2 100644
--- a/drivers/net/ethernet/qualcomm/qca_7k.h
+++ b/drivers/net/ethernet/qualcomm/qca_7k.h
@@ -67,6 +67,5 @@
void qcaspi_spi_error(struct qcaspi *qca);
int qcaspi_read_register(struct qcaspi *qca, u16 reg, u16 *result);
int qcaspi_write_register(struct qcaspi *qca, u16 reg, u16 value);
-int qcaspi_tx_cmd(struct qcaspi *qca, u16 cmd);
#endif /* _QCA_7K_H */
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
index a239ac4..c727357 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.c
+++ b/drivers/net/ethernet/qualcomm/qca_spi.c
@@ -192,6 +192,30 @@ qcaspi_read_legacy(struct qcaspi *qca, u8 *dst, u32 len)
}
static int
+qcaspi_tx_cmd(struct qcaspi *qca, u16 cmd)
+{
+ __be16 tx_data;
+ struct spi_message *msg = &qca->spi_msg1;
+ struct spi_transfer *transfer = &qca->spi_xfer1;
+ int ret;
+
+ tx_data = cpu_to_be16(cmd);
+ transfer->len = sizeof(tx_data);
+ transfer->tx_buf = &tx_data;
+ transfer->rx_buf = NULL;
+
+ ret = spi_sync(qca->spi_dev, msg);
+
+ if (!ret)
+ ret = msg->status;
+
+ if (ret)
+ qcaspi_spi_error(qca);
+
+ return ret;
+}
+
+static int
qcaspi_tx_frame(struct qcaspi *qca, struct sk_buff *skb)
{
u32 count;
--
2.1.4
^ permalink raw reply related
* [PATCH v5 06/17] net: qca_spi: remove QCASPI_MTU
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
There is no need for an additional MTU define.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_spi.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
index a0dbb92..a239ac4 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.c
+++ b/drivers/net/ethernet/qualcomm/qca_spi.c
@@ -69,7 +69,6 @@ static int qcaspi_pluggable = QCASPI_PLUGGABLE_MIN;
module_param(qcaspi_pluggable, int, 0);
MODULE_PARM_DESC(qcaspi_pluggable, "Pluggable SPI connection (yes/no).");
-#define QCASPI_MTU QCAFRM_MAX_MTU
#define QCASPI_TX_TIMEOUT (1 * HZ)
#define QCASPI_QCA7K_REBOOT_TIME_MS 1000
@@ -745,7 +744,7 @@ qcaspi_netdev_init(struct net_device *dev)
{
struct qcaspi *qca = netdev_priv(dev);
- dev->mtu = QCASPI_MTU;
+ dev->mtu = QCAFRM_MAX_MTU;
dev->type = ARPHRD_ETHER;
qca->clkspeed = qcaspi_clkspeed;
qca->burst_len = qcaspi_burst_len;
--
2.1.4
^ permalink raw reply related
* [PATCH v5 05/17] net: qualcomm: Improve readability of length defines
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
In order to avoid mixing things up, make the MTU and frame length
defines easier to read.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_framing.c | 2 +-
drivers/net/ethernet/qualcomm/qca_framing.h | 8 ++++----
drivers/net/ethernet/qualcomm/qca_spi.c | 12 ++++++------
3 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_framing.c b/drivers/net/ethernet/qualcomm/qca_framing.c
index faa924c..2341f2b 100644
--- a/drivers/net/ethernet/qualcomm/qca_framing.c
+++ b/drivers/net/ethernet/qualcomm/qca_framing.c
@@ -117,7 +117,7 @@ qcafrm_fsm_decode(struct qcafrm_handle *handle, u8 *buf, u16 buf_len, u8 recv_by
break;
case QCAFRM_WAIT_RSVD_BYTE2:
len = handle->offset;
- if (len > buf_len || len < QCAFRM_ETHMINLEN) {
+ if (len > buf_len || len < QCAFRM_MIN_LEN) {
ret = QCAFRM_INVLEN;
handle->state = QCAFRM_HW_LEN0;
} else {
diff --git a/drivers/net/ethernet/qualcomm/qca_framing.h b/drivers/net/ethernet/qualcomm/qca_framing.h
index 8b385e6..5df7c65 100644
--- a/drivers/net/ethernet/qualcomm/qca_framing.h
+++ b/drivers/net/ethernet/qualcomm/qca_framing.h
@@ -44,12 +44,12 @@
#define QCAFRM_INVFRAME (QCAFRM_ERR_BASE - 4)
/* Min/Max Ethernet MTU: 46/1500 */
-#define QCAFRM_ETHMINMTU (ETH_ZLEN - ETH_HLEN)
-#define QCAFRM_ETHMAXMTU ETH_DATA_LEN
+#define QCAFRM_MIN_MTU (ETH_ZLEN - ETH_HLEN)
+#define QCAFRM_MAX_MTU ETH_DATA_LEN
/* Min/Max frame lengths */
-#define QCAFRM_ETHMINLEN (QCAFRM_ETHMINMTU + ETH_HLEN)
-#define QCAFRM_ETHMAXLEN (QCAFRM_ETHMAXMTU + VLAN_ETH_HLEN)
+#define QCAFRM_MIN_LEN (QCAFRM_MIN_MTU + ETH_HLEN)
+#define QCAFRM_MAX_LEN (QCAFRM_MAX_MTU + VLAN_ETH_HLEN)
/* QCA7K header len */
#define QCAFRM_HEADER_LEN 8
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
index 5c79612..a0dbb92 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.c
+++ b/drivers/net/ethernet/qualcomm/qca_spi.c
@@ -69,7 +69,7 @@ static int qcaspi_pluggable = QCASPI_PLUGGABLE_MIN;
module_param(qcaspi_pluggable, int, 0);
MODULE_PARM_DESC(qcaspi_pluggable, "Pluggable SPI connection (yes/no).");
-#define QCASPI_MTU QCAFRM_ETHMAXMTU
+#define QCASPI_MTU QCAFRM_MAX_MTU
#define QCASPI_TX_TIMEOUT (1 * HZ)
#define QCASPI_QCA7K_REBOOT_TIME_MS 1000
@@ -402,7 +402,7 @@ qcaspi_tx_ring_has_space(struct tx_ring *txr)
if (txr->skb[txr->tail])
return 0;
- return (txr->size + QCAFRM_ETHMAXLEN < QCASPI_HW_BUF_LEN) ? 1 : 0;
+ return (txr->size + QCAFRM_MAX_LEN < QCASPI_HW_BUF_LEN) ? 1 : 0;
}
/* Flush the tx ring. This function is only safe to
@@ -666,8 +666,8 @@ qcaspi_netdev_xmit(struct sk_buff *skb, struct net_device *dev)
struct sk_buff *tskb;
u8 pad_len = 0;
- if (skb->len < QCAFRM_ETHMINLEN)
- pad_len = QCAFRM_ETHMINLEN - skb->len;
+ if (skb->len < QCAFRM_MIN_LEN)
+ pad_len = QCAFRM_MIN_LEN - skb->len;
if (qca->txr.skb[qca->txr.tail]) {
netdev_warn(qca->net_dev, "queue was unexpectedly full!\n");
@@ -804,8 +804,8 @@ qcaspi_netdev_setup(struct net_device *dev)
dev->tx_queue_len = 100;
/* MTU range: 46 - 1500 */
- dev->min_mtu = QCAFRM_ETHMINMTU;
- dev->max_mtu = QCAFRM_ETHMAXMTU;
+ dev->min_mtu = QCAFRM_MIN_MTU;
+ dev->max_mtu = QCAFRM_MAX_MTU;
qca = netdev_priv(dev);
memset(qca, 0, sizeof(struct qcaspi));
--
2.1.4
^ permalink raw reply related
* [PATCH v5 04/17] net: qualcomm: use net_device_ops instead of direct call
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby,
linux-serial-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
netdev-u79uwXL29TY76Z2rM5mHXA, devicetree-u79uwXL29TY76Z2rM5mHXA,
Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren-eS4NqCHxEME@public.gmane.org>
There is no need to export qcaspi_netdev_open and qcaspi_netdev_close
because they are also accessible via the net_device_ops.
Signed-off-by: Stefan Wahren <stefan.wahren-eS4NqCHxEME@public.gmane.org>
---
drivers/net/ethernet/qualcomm/qca_debug.c | 5 +++--
drivers/net/ethernet/qualcomm/qca_spi.c | 4 ++--
drivers/net/ethernet/qualcomm/qca_spi.h | 3 ---
3 files changed, 5 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_debug.c b/drivers/net/ethernet/qualcomm/qca_debug.c
index d145df9..92b6be9 100644
--- a/drivers/net/ethernet/qualcomm/qca_debug.c
+++ b/drivers/net/ethernet/qualcomm/qca_debug.c
@@ -275,6 +275,7 @@ qcaspi_get_ringparam(struct net_device *dev, struct ethtool_ringparam *ring)
static int
qcaspi_set_ringparam(struct net_device *dev, struct ethtool_ringparam *ring)
{
+ const struct net_device_ops *ops = dev->netdev_ops;
struct qcaspi *qca = netdev_priv(dev);
if ((ring->rx_pending) ||
@@ -283,13 +284,13 @@ qcaspi_set_ringparam(struct net_device *dev, struct ethtool_ringparam *ring)
return -EINVAL;
if (netif_running(dev))
- qcaspi_netdev_close(dev);
+ ops->ndo_stop(dev);
qca->txr.count = max_t(u32, ring->tx_pending, TX_RING_MIN_LEN);
qca->txr.count = min_t(u16, qca->txr.count, TX_RING_MAX_LEN);
if (netif_running(dev))
- qcaspi_netdev_open(dev);
+ ops->ndo_open(dev);
return 0;
}
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
index 8590109..5c79612 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.c
+++ b/drivers/net/ethernet/qualcomm/qca_spi.c
@@ -602,7 +602,7 @@ qcaspi_intr_handler(int irq, void *data)
return IRQ_HANDLED;
}
-int
+static int
qcaspi_netdev_open(struct net_device *dev)
{
struct qcaspi *qca = netdev_priv(dev);
@@ -639,7 +639,7 @@ qcaspi_netdev_open(struct net_device *dev)
return 0;
}
-int
+static int
qcaspi_netdev_close(struct net_device *dev)
{
struct qcaspi *qca = netdev_priv(dev);
diff --git a/drivers/net/ethernet/qualcomm/qca_spi.h b/drivers/net/ethernet/qualcomm/qca_spi.h
index 6e31a0e..064853d 100644
--- a/drivers/net/ethernet/qualcomm/qca_spi.h
+++ b/drivers/net/ethernet/qualcomm/qca_spi.h
@@ -108,7 +108,4 @@ struct qcaspi {
u16 burst_len;
};
-int qcaspi_netdev_open(struct net_device *dev);
-int qcaspi_netdev_close(struct net_device *dev);
-
#endif /* _QCA_SPI_H */
--
2.1.4
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH v5 03/17] net: qca_7k: Use BIT macro
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
Use the BIT macro for the CONFIG and INT register values.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_7k.h | 14 +++++++-------
1 file changed, 7 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_7k.h b/drivers/net/ethernet/qualcomm/qca_7k.h
index 1cad851..4047f0a 100644
--- a/drivers/net/ethernet/qualcomm/qca_7k.h
+++ b/drivers/net/ethernet/qualcomm/qca_7k.h
@@ -54,15 +54,15 @@
#define SPI_REG_ACTION_CTRL 0x1B00
/* SPI_CONFIG register definition; */
-#define QCASPI_SLAVE_RESET_BIT (1 << 6)
+#define QCASPI_SLAVE_RESET_BIT BIT(6)
/* INTR_CAUSE/ENABLE register definition. */
-#define SPI_INT_WRBUF_BELOW_WM (1 << 10)
-#define SPI_INT_CPU_ON (1 << 6)
-#define SPI_INT_ADDR_ERR (1 << 3)
-#define SPI_INT_WRBUF_ERR (1 << 2)
-#define SPI_INT_RDBUF_ERR (1 << 1)
-#define SPI_INT_PKT_AVLBL (1 << 0)
+#define SPI_INT_WRBUF_BELOW_WM BIT(10)
+#define SPI_INT_CPU_ON BIT(6)
+#define SPI_INT_ADDR_ERR BIT(3)
+#define SPI_INT_WRBUF_ERR BIT(2)
+#define SPI_INT_RDBUF_ERR BIT(1)
+#define SPI_INT_PKT_AVLBL BIT(0)
void qcaspi_spi_error(struct qcaspi *qca);
int qcaspi_read_register(struct qcaspi *qca, u16 reg, u16 *result);
--
2.1.4
^ permalink raw reply related
* [PATCH v5 02/17] net: qca_framing: use u16 for frame offset
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
It doesn't make sense to use a signed variable for offset here, so
fix it up.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_framing.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_framing.h b/drivers/net/ethernet/qualcomm/qca_framing.h
index d5e795d..8b385e6 100644
--- a/drivers/net/ethernet/qualcomm/qca_framing.h
+++ b/drivers/net/ethernet/qualcomm/qca_framing.h
@@ -103,7 +103,7 @@ struct qcafrm_handle {
enum qcafrm_state state;
/* Offset in buffer (borrowed for length too) */
- s16 offset;
+ u16 offset;
/* Frame length as kept by this module */
u16 len;
--
2.1.4
^ permalink raw reply related
* [PATCH v5 01/17] net: qualcomm: remove unnecessary includes
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
In-Reply-To: <1494406408-31760-1-git-send-email-stefan.wahren@i2se.com>
Most of the includes in qca_7k.c are unnecessary so we better remove them.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
drivers/net/ethernet/qualcomm/qca_7k.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/qca_7k.c b/drivers/net/ethernet/qualcomm/qca_7k.c
index f0066fb..557d53c 100644
--- a/drivers/net/ethernet/qualcomm/qca_7k.c
+++ b/drivers/net/ethernet/qualcomm/qca_7k.c
@@ -23,11 +23,7 @@
* kernel-based SPI device.
*/
-#include <linux/init.h>
-#include <linux/module.h>
-#include <linux/moduleparam.h>
#include <linux/spi/spi.h>
-#include <linux/version.h>
#include "qca_7k.h"
--
2.1.4
^ permalink raw reply related
* [PATCH v5 00/17] net: qualcomm: add QCA7000 UART driver
From: Stefan Wahren @ 2017-05-10 8:53 UTC (permalink / raw)
To: Rob Herring, David S. Miller
Cc: Mark Rutland, Greg Kroah-Hartman, Jiri Slaby, linux-serial,
linux-kernel, netdev, devicetree, Stefan Wahren
The Qualcomm QCA7000 HomePlug GreenPHY supports two interfaces:
UART and SPI. This patch series adds the missing support for UART.
This driver based on the Qualcomm code [1], but contains some changes:
* use random MAC address per default
* use net_device_stats from device
* share frame decoding between SPI and UART driver
* improve error handling
* reimplement tty_wakeup with work queue (based on slcan)
* use new serial device bus instead of ldisc
The patches 1 - 3 are just for clean up and are not related to
the UART support. Patches 4 - 15 prepare the existing QCA7000
code for UART support. Patch 16 is a improvement for serial device
bus. Patch 17 contains the new driver.
Cherry picking of the dt-bindings and serdev patch is okay. The
UART driver functionally (not compile) depends on this unmerged
patch [2].
The code itself has been tested on a Freescale i.MX28 board and
a Raspberry Pi Zero.
Changes in v5:
* rebase to current linux-next
* fix alignment issues in rx path
* fix buffer overrun with big ethernet frames
* fix init of UART decoding fsm
* add device UART settings to Kconfig help
* add current-speed to slave-device binding
* merge SPI and UART binding document
* rename qca_common to qca_7k_common
* drop patch for retrieving UART settings
* several cleanups
Changes in v4:
* rebase to current linux-next
* use parameter -M for git format-patch
* change order of local variables where possible
* implement basic serdev support (without hardware flow control)
Changes in v3:
* rebase to current net-next
Changes in v2:
* fix build issue by using netif_trans_update() and dev_trans_start()
[1] - https://github.com/IoE/qca7000
[2] - http://marc.info/?l=linux-serial&m=149338017301309&w=2
Stefan Wahren (17):
net: qualcomm: remove unnecessary includes
net: qca_framing: use u16 for frame offset
net: qca_7k: Use BIT macro
net: qualcomm: use net_device_ops instead of direct call
net: qualcomm: Improve readability of length defines
net: qca_spi: remove QCASPI_MTU
net: qualcomm: move qcaspi_tx_cmd to qca_spi.c
net: qca_spi: Clarify MODULE_DESCRIPTION
net: qualcomm: rename qca_framing.c to qca_7k_common.c
net: qualcomm: prepare frame decoding for UART driver
net: qualcomm: make qca_7k_common a separate kernel module
dt-bindings: qca7000-spi: Rework binding
dt-bindings: qca7000: rename binding
dt-bindings: slave-device: add current-speed property
dt-bindings: qca7000: append UART interface to binding
tty: serdev-ttyport: return actual baudrate from ttyport_set_baudrate
net: qualcomm: add QCA7000 UART driver
.../devicetree/bindings/net/qca-qca7000-spi.txt | 47 ---
.../devicetree/bindings/net/qca-qca7000.txt | 88 +++++
.../devicetree/bindings/serial/slave-device.txt | 9 +
drivers/net/ethernet/qualcomm/Kconfig | 24 +-
drivers/net/ethernet/qualcomm/Makefile | 7 +-
drivers/net/ethernet/qualcomm/qca_7k.c | 28 --
drivers/net/ethernet/qualcomm/qca_7k.h | 15 +-
.../qualcomm/{qca_framing.c => qca_7k_common.c} | 26 +-
.../qualcomm/{qca_framing.h => qca_7k_common.h} | 24 +-
drivers/net/ethernet/qualcomm/qca_debug.c | 5 +-
drivers/net/ethernet/qualcomm/qca_spi.c | 47 ++-
drivers/net/ethernet/qualcomm/qca_spi.h | 5 +-
drivers/net/ethernet/qualcomm/qca_uart.c | 423 +++++++++++++++++++++
drivers/tty/serdev/serdev-ttyport.c | 2 +-
14 files changed, 630 insertions(+), 120 deletions(-)
delete mode 100644 Documentation/devicetree/bindings/net/qca-qca7000-spi.txt
create mode 100644 Documentation/devicetree/bindings/net/qca-qca7000.txt
rename drivers/net/ethernet/qualcomm/{qca_framing.c => qca_7k_common.c} (85%)
rename drivers/net/ethernet/qualcomm/{qca_framing.h => qca_7k_common.h} (86%)
create mode 100644 drivers/net/ethernet/qualcomm/qca_uart.c
--
2.1.4
^ permalink raw reply
* RE: [PATCH 1/4] hamradio: Combine two seq_printf() calls into one in yam_seq_show()
From: David Laight @ 2017-05-10 8:48 UTC (permalink / raw)
To: 'SF Markus Elfring', linux-hams@vger.kernel.org,
netdev@vger.kernel.org, David S. Miller, Javier Martinez Canillas,
Jean-Paul Roubelat
Cc: LKML, kernel-janitors@vger.kernel.org
In-Reply-To: <c2866b46-28c9-f17f-dce7-96425a17a43a@users.sourceforge.net>
From: SF Markus Elfring
> Sent: 09 May 2017 15:22
> A bit of data was put into a sequence by two separate function calls.
> Print the same data by a single function call instead.
>
> This issue was detected by using the Coccinelle software.
>
> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
> ---
> drivers/net/hamradio/yam.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/net/hamradio/yam.c b/drivers/net/hamradio/yam.c
> index b6891ada1d7b..542f1e511df1 100644
> --- a/drivers/net/hamradio/yam.c
> +++ b/drivers/net/hamradio/yam.c
> @@ -830,8 +830,7 @@ static int yam_seq_show(struct seq_file *seq, void *v)
> seq_printf(seq, " RxFrames %lu\n", dev->stats.rx_packets);
> seq_printf(seq, " TxInt %u\n", yp->nb_mdint);
> seq_printf(seq, " RxInt %u\n", yp->nb_rxint);
> - seq_printf(seq, " RxOver %lu\n", dev->stats.rx_fifo_errors);
> - seq_printf(seq, "\n");
> + seq_printf(seq, " RxOver %lu\n\n", dev->stats.rx_fifo_errors);
> return 0;
The code was consistently (and probably deliberately) using one seq_printf()
call for each line so that the source looks 'a bit like' the output.
These changes are all stupid.
David
^ permalink raw reply
* Re: [Patch net] ipv4: restore rt->fi for reference counting
From: Julian Anastasov @ 2017-05-10 7:38 UTC (permalink / raw)
To: Cong Wang
Cc: Eric Dumazet, David Miller, Linux Kernel Network Developers,
Andrey Konovalov, Eric Dumazet
In-Reply-To: <CAM_iQpU+fXO7eFroAYMv6vqDzCK_ZYXjrTPAMfoDR2BDqaK9rQ@mail.gmail.com>
Hello,
On Tue, 9 May 2017, Cong Wang wrote:
> > Also setting nexthop_nh->nh_dev to NULL looks quite dangerous
> >
> > We have plenty of sites doing :
> >
> > if (fi->fib_dev)
> > x = fi->fib_dev->field
> >
> > fib_route_seq_show() is one example.
> >
>
> All of them take RCU read lock, so, as I explained in the code comment,
> they all should be fine because of synchronize_net() on unregister path.
> Do you see anything otherwise?
During NETDEV_UNREGISTER packets for dev should not
be flying but packets for other devs can walk the nexthops
for multipath routes. It is the rcu_barrier before
NETDEV_UNREGISTER_FINAL that allows nh_dev to be set to NULL
during this grace period but there are many places to fix that
assume nh_dev!=NULL.
But why we leak routes? Is there some place that holds
routes without listening for NETDEV_UNREGISTER? On fib_flush
the infos are unlinked from trees, so after a grace period packets
should not see/hold such infos. If we hold routes somewhere for
long time, problem can happen also for routes with single nexthop.
Regards
^ permalink raw reply
* Re: [PATCH iproute2 v2 1/1] vxlan: Add support for modifying vxlan device attributes
From: Girish Moodalbail @ 2017-05-10 7:31 UTC (permalink / raw)
To: Roman Mashak; +Cc: stephen, netdev
In-Reply-To: <85shkgd92a.fsf@mojatatu.com>
On 5/7/17 7:40 AM, Roman Mashak wrote:
> Girish Moodalbail <girish.moodalbail@oracle.com> writes:
>
>
> [...]
>
>> ip/iplink_vxlan.c | 251 +++++++++++++++++++++++++++++++-----------------------
>> 1 file changed, 143 insertions(+), 108 deletions(-)
>>
>> diff --git a/ip/iplink_vxlan.c b/ip/iplink_vxlan.c
>> index b4ebb13..2bd619d 100644
>> --- a/ip/iplink_vxlan.c
>> +++ b/ip/iplink_vxlan.c
>> @@ -21,6 +21,8 @@
>> #include "utils.h"
>> #include "ip_common.h"
>>
>> +#define VXLAN_ATTRSET(attrs, type) (((attrs) & (1L << (type))) != 0)
>
> I think you can drop '!= 0' part in the macro.
Sure it can be done that way as well. However, I have seen both forms in use in
iproute2 code base, so I would like to keep this as is unless you feel strongly
about it. Furthermore, running 'checkpatch.pl --strict' on the patch didn't
complain too.
regards,
~Girish
>
> [...]
>
>>
>> +static void check_duparg(__u64 *attrs, int type, const char *key,
>> + const char *argv)
>> +{
>> + if (!VXLAN_ATTRSET(*attrs, type)) {
>> + *attrs |= (1L << type);
>> + return;
>> + }
>> + duparg2(key, argv);
>> +}
>
> [...]
>
^ permalink raw reply
* Re: [PATCH] wcn36xx: Close SMD channel on device removal
From: Arend van Spriel @ 2017-05-10 7:27 UTC (permalink / raw)
To: Bjorn Andersson, Kalle Valo
Cc: Eugene Krasnikov, Eyal Ilsar, wcn36xx, linux-wireless, netdev,
linux-kernel, linux-arm-msm
In-Reply-To: <20170509230312.GR15143@minitux>
On 5/10/2017 1:03 AM, Bjorn Andersson wrote:
> On Mon 08 May 23:17 PDT 2017, Kalle Valo wrote:
>
>> Bjorn Andersson <bjorn.andersson@linaro.org> writes:
>>
>>> The SMD channel is not the primary WCNSS channel and must explicitly be
>>> closed as the device is removed, or the channel will already by open on
>>> a subsequent probe call in e.g. the case of reloading the kernel module.
>>>
>>> This issue was introduced because I simplified the underlying SMD
>>> implementation while the SMD adaptions of the driver sat on the mailing
>>> list, but missed to update these patches. The patch does however only
>>> apply back to the transition to rpmsg, hence the limited Fixes.
>>>
>>> Fixes: 5052de8deff5 ("soc: qcom: smd: Transition client drivers from smd to rpmsg")
>>> Reported-by: Eyal Ilsar <c_eilsar@qti.qualcomm.com>
>>> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
>>
>> As this is a regression I'll queue this to 4.12.
>>
>
> Thanks.
>
>> But if this is an older bug (didn't quite understand your description
>> though) should there be a separate patch for stable releases?
>>
>
> AFAICT this never worked, as it seems I did the rework in SMD while we
> tried to figure out the dependency issues we had with moving to SMD. So
> v4.9 through v4.11 has SMD support - with this bug.
>
> How do I proceed, do you want me to write up a fix for stable@? Do I
> send that out as an ordinary patch?
If the patch applies cleanly on branches linux-4.9.y through
linux-4.11.y in the stable repository you can go for '--- Option 1 ---'
as described in <linux-repo>/Documentation/stable_kernel_rules.txt.
Regards,
Arend
^ permalink raw reply
* [PATCH] net/smc: mark as BROKEN due to remote memory exposure
From: Christoph Hellwig @ 2017-05-10 7:26 UTC (permalink / raw)
To: davem; +Cc: ubraun, netdev, linux-rdma, stable
The driver has a lot of quality issues due to the lack of RDMA-side
review, and explicitly bypasses APIs to register all memory once a
connection is made, and thus allows remote access to memoery.
Mark it as broken until at least that part is fixed.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org
---
net/smc/Kconfig | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/net/smc/Kconfig b/net/smc/Kconfig
index c717ef0896aa..fe6b78bc515f 100644
--- a/net/smc/Kconfig
+++ b/net/smc/Kconfig
@@ -1,6 +1,6 @@
config SMC
tristate "SMC socket protocol family"
- depends on INET && INFINIBAND
+ depends on INET && INFINIBAND && BROKEN
---help---
SMC-R provides a "sockets over RDMA" solution making use of
RDMA over Converged Ethernet (RoCE) technology to upgrade
@@ -8,6 +8,10 @@ config SMC
The Linux implementation of the SMC-R solution is designed as
a separate socket family SMC.
+ Warning: SMC will expose all memory for remote reads and writes
+ once a connection is established. Don't enable this option except
+ for tightly controlled lab environment.
+
Select this option if you want to run SMC socket applications
config SMC_DIAG
--
2.11.0
^ permalink raw reply related
* (unknown),
From: kelley @ 2017-05-10 7:23 UTC (permalink / raw)
To: netdev
[-- Attachment #1: 620_netdev.zip --]
[-- Type: application/zip, Size: 1923 bytes --]
^ permalink raw reply
* Re: [PATCH] net: ethernet: stmmac: properly set PS bit in MII configurations during reset
From: Thomas Petazzoni @ 2017-05-10 7:18 UTC (permalink / raw)
To: Giuseppe CAVALLARO; +Cc: Corentin Labbe, Alexandre Torgue, netdev, stable
In-Reply-To: <e8151650-aa51-6713-f10c-b3663763ff78@st.com>
Hello,
On Wed, 10 May 2017 09:03:12 +0200, Giuseppe CAVALLARO wrote:
> > Please, read again my patch and the description of the problem that I
> > have sent. But basically, any solution that does not allow to set the
> > PS bit between asserting the DMA reset bit and polling for it to clear
> > will not work for MII PHYs.
>
> yes your point was clear to me, I was just wondering if we could find an
> easier way
> to solve it w/o changing the API, adding the set_ps and propagating the
> "interface"
> inside the DMA reset.
>
> Maybe this could be fixed in the glue-logic in some way. Let me know
> what do you think.
Well, it's more up to you to tell me how you would like this be solved.
We figured out what the problem was, but I don't know well enough the
architecture of the driver to decide how the solution to this problem
should be designed. I made an initial simple proposal to show what is
needed, but I'm definitely open to suggestions.
Best regards,
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
^ permalink raw reply
* Re: [PATCH] net: ethernet: stmmac: properly set PS bit in MII configurations during reset
From: Giuseppe CAVALLARO @ 2017-05-10 7:03 UTC (permalink / raw)
To: Thomas Petazzoni; +Cc: Corentin Labbe, Alexandre Torgue, netdev, stable
In-Reply-To: <20170508211230.58aeead9@free-electrons.com>
Hi Thomas
On 5/8/2017 9:12 PM, Thomas Petazzoni wrote:
> Hello,
>
> On Mon, 8 May 2017 16:28:21 +0200, Giuseppe CAVALLARO wrote:
>
>>> I just see that GMAC_CONTROL and MAC_CTRL_REG are the same, so why not create a custom adjust_link for each dwmac type ?
>>> This will permit to call it instead of set_ps() and remove lots of if (has_gmac) and co in stmmac_adjust_link()
>>> Basicly replace all between "ctrl = readl()... and writel(ctrl)" by a sot of priv->hw->mac->adjust_link()
>>>
>>> It will also help a lot for my dwmac-sun8i inclusion (since I add some if has_sun8i:))
>> Corentin, I think this is a good idea and maybe necessary now that the
>> driver is supporting a lot of chips.
>> In the past it was sufficient to have a adjust link function and a
>> stmmac_hw_fix_mac_speed
>> to invoke dedicated hook shared between MAC10/100 and GMAC inside STM
>> platforms.
>>
>> Thomas, I wonder if you could take a look at the
>> priv->plat->fix_mac_speed. This can be used
>> for setting internal registers too.
> Once again, this is not called at the right time to fix the issue I'm
> seeing with a MII PHY. I need to adjust the PS bit between asserting the
> reset and polling for the reset bit to clear.
>
> ->fix_mac_speed() is called in the adjust_link() call-back, which is
> called way too late.
>
> Please, read again my patch and the description of the problem that I
> have sent. But basically, any solution that does not allow to set the
> PS bit between asserting the DMA reset bit and polling for it to clear
> will not work for MII PHYs.
yes your point was clear to me, I was just wondering if we could find an
easier way
to solve it w/o changing the API, adding the set_ps and propagating the
"interface"
inside the DMA reset.
Maybe this could be fixed in the glue-logic in some way. Let me know
what do you think.
peppe
>
> Best regards,
>
> Thomas Petazzoni
^ permalink raw reply
* Re: [PATCH 0/4] net: stmmac: Fine-tuning for four function implementations
From: Giuseppe CAVALLARO @ 2017-05-10 6:56 UTC (permalink / raw)
To: SF Markus Elfring, netdev, Alexandre Torgue; +Cc: LKML, kernel-janitors
In-Reply-To: <aa58027a-a39c-963e-4376-a4d5312ee118@users.sourceforge.net>
Hello Markus
Thanks a lot for your effort on stmmac
On 5/9/2017 1:51 PM, SF Markus Elfring wrote:
> From: Markus Elfring <elfring@users.sourceforge.net>
> Date: Tue, 9 May 2017 13:48:03 +0200
>
> A few update suggestions were taken into account
> from static source code analysis.
>
> Markus Elfring (4):
> Combine three seq_printf() calls into a seq_puts() in stmmac_sysfs_dma_cap_read()
> Replace five seq_printf() calls by seq_puts()
> Use seq_putc() in sysfs_display_ring()
> Delete an unnecessary return statement in stmmac_get_tx_hwtstamp()
>
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 22 ++++++++++------------
> 1 file changed, 10 insertions(+), 12 deletions(-)
Acked-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
^ permalink raw reply
* Re: bpf pointer alignment validation
From: Alexei Starovoitov @ 2017-05-10 5:57 UTC (permalink / raw)
To: David Miller; +Cc: daniel, ast, netdev
In-Reply-To: <20170509.143234.1785658771452710730.davem@davemloft.net>
On Tue, May 09, 2017 at 02:32:34PM -0400, David Miller wrote:
>
> +static u32 calc_align(u32 imm)
> +{
> + u32 align = 1;
> +
> + if (!imm)
> + return 1U << 31;
> +
> + while (!(imm & 1)) {
> + imm >>= 1;
> + align <<= 1;
> + }
> + return align;
> +}
same question as in previous reply.
Why not to use something like:
static u32 calc_align(u32 n)
{
if (!n)
return 1U << 31;
return n - ((n - 1) & n);
}
> - if (log_level && do_print_state) {
> + if (log_level > 1 || (log_level && do_print_state)) {
> verbose("\nfrom %d to %d:", prev_insn_idx, insn_idx);
> print_verifier_state(&env->cur_state);
> do_print_state = false;
this needs to be tweaked like
if (log_level > 1)
verbose("%d:", insn_idx);
else
verbose("\nfrom %d to %d:", prev_insn_idx, insn_idx);
otherwise it prints prev_insn_idx which is meaningful
only with processing exit and search pruning.
That's why below ...
> + .descr = "unknown shift",
> + .insns = {
> + LOAD_UNKNOWN(BPF_REG_3),
> + BPF_ALU64_IMM(BPF_LSH, BPF_REG_3, 1),
> + BPF_ALU64_IMM(BPF_LSH, BPF_REG_3, 1),
> + BPF_ALU64_IMM(BPF_LSH, BPF_REG_3, 1),
> + BPF_ALU64_IMM(BPF_LSH, BPF_REG_3, 1),
> + LOAD_UNKNOWN(BPF_REG_4),
> + BPF_ALU64_IMM(BPF_LSH, BPF_REG_4, 5),
> + BPF_ALU64_IMM(BPF_RSH, BPF_REG_4, 1),
> + BPF_ALU64_IMM(BPF_RSH, BPF_REG_4, 1),
> + BPF_ALU64_IMM(BPF_RSH, BPF_REG_4, 1),
> + BPF_ALU64_IMM(BPF_RSH, BPF_REG_4, 1),
> + BPF_MOV64_IMM(BPF_REG_0, 0),
> + BPF_EXIT_INSN(),
> + },
> + .prog_type = BPF_PROG_TYPE_SCHED_CLS,
> + .matches = {
> + "from 4 to 7: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=inv56 R10=fp",
> + "from 4 to 8: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=inv55,min_align=2 R10=fp",
> + "from 4 to 9: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=inv54,min_align=4 R10=fp",
> + "from 4 to 10: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=inv53,min_align=8 R10=fp",
> + "from 4 to 11: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=inv52,min_align=16 R10=fp",
> + "from 15 to 18: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=inv56 R10=fp",
> + "from 15 to 19: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=inv51,min_align=32 R10=fp",
> + "from 15 to 20: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=inv52,min_align=16 R10=fp",
... looks crazy here, since program is linear and 'from 4' and 'from 15' are
completely non-obvious and will change in the future for sure.
Since it just happened that search pruning heuristic kicked in at those points.
Hence doing verbose("%d:", insn_idx); is necessary to avoid noise.
> + .prog_type = BPF_PROG_TYPE_SCHED_CLS,
> + .matches = {
> + /* Calculated offset in R4 has unknown value, but known
> + * alignment of 4.
> + */
> + "from 4 to 8: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R6=inv54,min_align=4 R10=fp",
> +
> + /* Offset is added to packet pointer, resulting in known
> + * auxiliary alignment and offset.
> + */
> + "from 4 to 11: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R5=pkt(id=1,off=0,r=0),aux_off=14,aux_off_align=4 R6=inv54,min_align=4 R10=fp",
> +
> + /* At the time the word size load is performed from R5,
> + * it's total offset is NET_IP_ALIGN + reg->off (0) +
> + * reg->aux_off (14) which is 16. Then the variable
> + * offset is considered using reg->aux_off_align which
> + * is 4 and meets the load's requirements.
> + */
> + "from 13 to 15: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=pkt(id=1,off=4,r=4),aux_off=14,aux_off_align=4 R5=pkt(id=1,off=0,r=4),aux_off=14,aux_off_align=4 R6=inv54,min_align=4 R10=fp",
> +
> +
> + /* Variable offset is added to R5 packet pointer,
> + * resulting in auxiliary alignment of 4.
> + */
> + "from 13 to 18: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=inv,aux_off=14,aux_off_align=4 R5=pkt(id=2,off=0,r=0),aux_off_align=4 R6=inv54,min_align=4 R10=fp",
> +
> + /* Constant offset is added to R5, resulting in
> + * reg->off of 14.
> + */
> + "from 13 to 19: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=inv,aux_off=14,aux_off_align=4 R5=pkt(id=2,off=14,r=0),aux_off_align=4 R6=inv54,min_align=4 R10=fp",
> +
> + /* At the time the word size load is performed from R5,
> + * it's total offset is NET_IP_ALIGN + reg->off (14) which
> + * is 16. Then the variable offset is considered using
> + * reg->aux_off_align which is 4 and meets the load's
> + * requirements.
> + */
> + "from 21 to 23: R0=pkt(id=0,off=8,r=8) R1=ctx R2=pkt(id=0,off=0,r=8) R3=pkt_end R4=pkt(id=2,off=18,r=18),aux_off_align=4 R5=pkt(id=2,off=14,r=18),aux_off_align=4 R6=inv54,min_align=4 R10=fp",
Nice to see all these comments.
I wonder how we can make them automatic in the verifier.
Like if verifier can somehow hint the user in such human friendly way
about what is happening with the program.
Today that's the #1 problem. Most folks complaining
that verifier error messages are too hard to understand.
CTF should help. Proper data-flow analysis in the verifier should help too.
> +static int do_test_single(struct bpf_align_test *test)
> +{
> + struct bpf_insn *prog = test->insns;
> + int prog_type = test->prog_type;
> + int prog_len, i;
> + int fd_prog;
> + int ret;
> +
> + prog_len = probe_filter_length(prog);
> + fd_prog = bpf_verify_program(prog_type ? : BPF_PROG_TYPE_SOCKET_FILTER,
> + prog, prog_len, "GPL", 0,
> + bpf_vlog, sizeof(bpf_vlog));
> + if (fd_prog < 0) {
> + printf("Failed to load program.\n");
> + printf("%s", bpf_vlog);
> + ret = 1;
> + } else {
> + ret = 0;
> + for (i = 0; i < MAX_MATCHES; i++) {
> + const char *t, *m = test->matches[i];
> +
> + if (!m)
> + break;
> + t = strstr(bpf_vlog, m);
> + if (!t) {
> + printf("Failed to find match: %s\n", m);
> + ret = 1;
> + printf("%s", bpf_vlog);
> + break;
> + }
> + }
> + /* printf("%s", bpf_vlog); */
> + close(fd_prog);
would it make sense to bpf_prog_test_run() it here as well?
On x86 not much value, but on sparc we can somehow look for traps?
Is there some counter of unaligned traps that we can read and report
as error to user space after prog_test_run ?
These tests we cannot really run, since they don't do any load/store.
I mean more for some future tests. Or some sort of debug warn
that there were traps while bpf prog was executed, so the user
is alarmed and reports to us, since that would be a bug in verifier
align logic?
^ permalink raw reply
* Re: DQL and TCQ_F_CAN_BYPASS destroy performance under virtualizaiton (Was: "Re: net_sched strange in 4.11")
From: Anton Ivanov @ 2017-05-10 5:35 UTC (permalink / raw)
To: Jason Wang, Stefan Hajnoczi; +Cc: David S. Miller, netdev, Michael S. Tsirkin
In-Reply-To: <52801b17-d8af-eaba-3ecf-fa4495c352f5@redhat.com>
[snip]
> Virtio-net net does not support BQL. Before commit ea7735d97ba9
> ("virtio-net: move free_old_xmit_skbs"), it's even impossible to
> support that since we don't have tx interrupt for each packet. I
> haven't measured the impact of xmit_more, maybe I was wrong but I
> think it may help in some cases since it may improve the batching on
> host more or less.
Sorry, hit send too soon.
Impact of xmit more depends on your transport.
If, for example, you are using sendmmsg on the outer side which can
consume the bulked data "as is", the impact is quite significant. If
your transport does not support bulking, the fact there was bulking
earlier in the chain has little impact.
There is some, but not a lot.
[snip]
--
Anton R. Ivanov
Cambridgegreys Limited. Registered in England. Company Number 10273661
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox