* [PATCH net-next 1/8] s390/cio: Add new Operation Code OC3 to PNSO
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 2/8] s390/cio: Helper functions to read CSSID, IID, and CHID Julian Wiedmann
` (7 subsequent siblings)
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
Add support for operation code 3 (OC3) of the
Perform-Network-Subchannel-Operations (PNSO) function
of the Channel-Subsystem-Call (CHSC) instruction.
PNSO provides 2 operation codes:
OC0 - BRIDGE_INFO
OC3 - ADDR_INFO (new)
Extend the function calls to *pnso* to pass the OC and
add new response code 0108.
Support for OC3 is indicated by a flag in the css_general_characteristics.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com>
Reviewed-by: Vineeth Vijayan <vneethv@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
---
arch/s390/include/asm/ccwdev.h | 5 ++---
arch/s390/include/asm/chsc.h | 7 +++++++
arch/s390/include/asm/css_chars.h | 4 +++-
drivers/s390/cio/chsc.c | 11 ++++++-----
drivers/s390/cio/chsc.h | 6 ++----
drivers/s390/cio/device_ops.c | 8 ++++----
drivers/s390/net/qeth_l2_main.c | 13 ++++++++-----
7 files changed, 32 insertions(+), 22 deletions(-)
diff --git a/arch/s390/include/asm/ccwdev.h b/arch/s390/include/asm/ccwdev.h
index 3cfe1eb89838..9739a00e2190 100644
--- a/arch/s390/include/asm/ccwdev.h
+++ b/arch/s390/include/asm/ccwdev.h
@@ -238,7 +238,6 @@ extern void ccw_device_get_schid(struct ccw_device *, struct subchannel_id *);
struct channel_path_desc_fmt0 *ccw_device_get_chp_desc(struct ccw_device *, int);
u8 *ccw_device_get_util_str(struct ccw_device *cdev, int chp_idx);
int ccw_device_pnso(struct ccw_device *cdev,
- struct chsc_pnso_area *pnso_area,
- struct chsc_pnso_resume_token resume_token,
- int cnc);
+ struct chsc_pnso_area *pnso_area, u8 oc,
+ struct chsc_pnso_resume_token resume_token, int cnc);
#endif /* _S390_CCWDEV_H_ */
diff --git a/arch/s390/include/asm/chsc.h b/arch/s390/include/asm/chsc.h
index 36ce2d25a5fc..ae4d2549cd67 100644
--- a/arch/s390/include/asm/chsc.h
+++ b/arch/s390/include/asm/chsc.h
@@ -11,6 +11,13 @@
#include <uapi/asm/chsc.h>
+/**
+ * Operation codes for CHSC PNSO:
+ * PNSO_OC_NET_BRIDGE_INFO - only addresses that are visible to a bridgeport
+ * PNSO_OC_NET_ADDR_INFO - all addresses
+ */
+#define PNSO_OC_NET_BRIDGE_INFO 0
+#define PNSO_OC_NET_ADDR_INFO 3
/**
* struct chsc_pnso_naid_l2 - network address information descriptor
* @nit: Network interface token
diff --git a/arch/s390/include/asm/css_chars.h b/arch/s390/include/asm/css_chars.h
index 480bb02ccacd..638137d46c85 100644
--- a/arch/s390/include/asm/css_chars.h
+++ b/arch/s390/include/asm/css_chars.h
@@ -36,7 +36,9 @@ struct css_general_char {
u64 alt_ssi : 1; /* bit 108 */
u64 : 1;
u64 narf : 1; /* bit 110 */
- u64 : 12;
+ u64 : 5;
+ u64 enarf: 1; /* bit 116 */
+ u64 : 6;
u64 util_str : 1;/* bit 123 */
} __packed;
diff --git a/drivers/s390/cio/chsc.c b/drivers/s390/cio/chsc.c
index c314e9495c1b..8f764a295a51 100644
--- a/drivers/s390/cio/chsc.c
+++ b/drivers/s390/cio/chsc.c
@@ -65,6 +65,8 @@ int chsc_error_from_response(int response)
case 0x0100:
case 0x0102:
return -ENOMEM;
+ case 0x0108: /* "HW limit exceeded" for the op 0x003d */
+ return -EUSERS;
default:
return -EIO;
}
@@ -1340,6 +1342,7 @@ EXPORT_SYMBOL_GPL(chsc_scm_info);
* chsc_pnso() - Perform Network-Subchannel Operation
* @schid: id of the subchannel on which PNSO is performed
* @pnso_area: request and response block for the operation
+ * @oc: Operation Code
* @resume_token: resume token for multiblock response
* @cnc: Boolean change-notification control
*
@@ -1347,10 +1350,8 @@ EXPORT_SYMBOL_GPL(chsc_scm_info);
*
* Returns 0 on success.
*/
-int chsc_pnso(struct subchannel_id schid,
- struct chsc_pnso_area *pnso_area,
- struct chsc_pnso_resume_token resume_token,
- int cnc)
+int chsc_pnso(struct subchannel_id schid, struct chsc_pnso_area *pnso_area,
+ u8 oc, struct chsc_pnso_resume_token resume_token, int cnc)
{
memset(pnso_area, 0, sizeof(*pnso_area));
pnso_area->request.length = 0x0030;
@@ -1359,7 +1360,7 @@ int chsc_pnso(struct subchannel_id schid,
pnso_area->ssid = schid.ssid;
pnso_area->sch = schid.sch_no;
pnso_area->cssid = schid.cssid;
- pnso_area->oc = 0; /* Store-network-bridging-information list */
+ pnso_area->oc = oc;
pnso_area->resume_token = resume_token;
pnso_area->n = (cnc != 0);
if (chsc(pnso_area))
diff --git a/drivers/s390/cio/chsc.h b/drivers/s390/cio/chsc.h
index 7ecf7e4c402e..7416957ba9f4 100644
--- a/drivers/s390/cio/chsc.h
+++ b/drivers/s390/cio/chsc.h
@@ -205,10 +205,8 @@ struct chsc_scm_info {
int chsc_scm_info(struct chsc_scm_info *scm_area, u64 token);
-int chsc_pnso(struct subchannel_id schid,
- struct chsc_pnso_area *pnso_area,
- struct chsc_pnso_resume_token resume_token,
- int cnc);
+int chsc_pnso(struct subchannel_id schid, struct chsc_pnso_area *pnso_area,
+ u8 oc, struct chsc_pnso_resume_token resume_token, int cnc);
int __init chsc_get_cssid(int idx);
diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c
index 963fcc9054c6..cdf44f398957 100644
--- a/drivers/s390/cio/device_ops.c
+++ b/drivers/s390/cio/device_ops.c
@@ -714,6 +714,7 @@ EXPORT_SYMBOL_GPL(ccw_device_get_schid);
* ccw_device_pnso() - Perform Network-Subchannel Operation
* @cdev: device on which PNSO is performed
* @pnso_area: request and response block for the operation
+ * @oc: Operation Code
* @resume_token: resume token for multiblock response
* @cnc: Boolean change-notification control
*
@@ -722,14 +723,13 @@ EXPORT_SYMBOL_GPL(ccw_device_get_schid);
* Returns 0 on success.
*/
int ccw_device_pnso(struct ccw_device *cdev,
- struct chsc_pnso_area *pnso_area,
- struct chsc_pnso_resume_token resume_token,
- int cnc)
+ struct chsc_pnso_area *pnso_area, u8 oc,
+ struct chsc_pnso_resume_token resume_token, int cnc)
{
struct subchannel_id schid;
ccw_device_get_schid(cdev, &schid);
- return chsc_pnso(schid, pnso_area, resume_token, cnc);
+ return chsc_pnso(schid, pnso_area, oc, resume_token, cnc);
}
EXPORT_SYMBOL_GPL(ccw_device_pnso);
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 491578009f12..2ab130d5c42d 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -642,6 +642,7 @@ static void qeth_l2_set_rx_mode(struct net_device *dev)
/**
* qeth_l2_pnso() - perform network subchannel operation
* @card: qeth_card structure pointer
+ * @oc: Operation Code
* @cnc: Boolean Change-Notification Control
* @cb: Callback function will be executed for each element
* of the address list
@@ -652,7 +653,7 @@ static void qeth_l2_set_rx_mode(struct net_device *dev)
* control" is set, further changes in the address list will be reported
* via the IPA command.
*/
-static int qeth_l2_pnso(struct qeth_card *card, int cnc,
+static int qeth_l2_pnso(struct qeth_card *card, u8 oc, int cnc,
void (*cb)(void *priv, struct chsc_pnso_naid_l2 *entry),
void *priv)
{
@@ -663,13 +664,14 @@ static int qeth_l2_pnso(struct qeth_card *card, int cnc,
int i, size, elems;
int rc;
- QETH_CARD_TEXT(card, 2, "PNSO");
rr = (struct chsc_pnso_area *)get_zeroed_page(GFP_KERNEL);
if (rr == NULL)
return -ENOMEM;
do {
+ QETH_CARD_TEXT(card, 2, "PNSO");
/* on the first iteration, naihdr.resume_token will be zero */
- rc = ccw_device_pnso(ddev, rr, rr->naihdr.resume_token, cnc);
+ rc = ccw_device_pnso(ddev, rr, oc, rr->naihdr.resume_token,
+ cnc);
if (rc)
continue;
if (cb == NULL)
@@ -1578,11 +1580,12 @@ int qeth_bridgeport_an_set(struct qeth_card *card, int enable)
if (enable) {
qeth_bridge_emit_host_event(card, anev_reset, 0, NULL, NULL);
qeth_l2_set_pnso_mode(card, QETH_PNSO_BRIDGEPORT);
- rc = qeth_l2_pnso(card, 1, qeth_bridgeport_an_set_cb, card);
+ rc = qeth_l2_pnso(card, PNSO_OC_NET_BRIDGE_INFO, 1,
+ qeth_bridgeport_an_set_cb, card);
if (rc)
qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
} else {
- rc = qeth_l2_pnso(card, 0, NULL, NULL);
+ rc = qeth_l2_pnso(card, PNSO_OC_NET_BRIDGE_INFO, 0, NULL, NULL);
qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
}
return rc;
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH net-next 2/8] s390/cio: Helper functions to read CSSID, IID, and CHID
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 1/8] s390/cio: Add new Operation Code OC3 to PNSO Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 3/8] s390/qeth: Detect PNSO OC3 capability Julian Wiedmann
` (6 subsequent siblings)
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
Add helper functions to expose Channel Subsystem ID (CSSID), MIF Image Id
(IID), Channel ID (CHID) and Channel Path ID (CHPID).
These values are required by the qeth driver's exploitation of network-
address-change-notifications to determine which entries belong to this
interface.
Store the Partition identifier in System log, as this may be used to map
a Linux view to a Hardware view for debugging purpose.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Vineeth Vijayan <vneethv@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
---
arch/s390/include/asm/ccwdev.h | 4 ++
drivers/s390/cio/chsc.c | 11 +++--
drivers/s390/cio/chsc.h | 2 +-
drivers/s390/cio/css.c | 11 +++--
drivers/s390/cio/css.h | 4 +-
drivers/s390/cio/device_ops.c | 85 ++++++++++++++++++++++++++++++++++
6 files changed, 108 insertions(+), 9 deletions(-)
diff --git a/arch/s390/include/asm/ccwdev.h b/arch/s390/include/asm/ccwdev.h
index 9739a00e2190..c0be5fe1ddba 100644
--- a/arch/s390/include/asm/ccwdev.h
+++ b/arch/s390/include/asm/ccwdev.h
@@ -240,4 +240,8 @@ u8 *ccw_device_get_util_str(struct ccw_device *cdev, int chp_idx);
int ccw_device_pnso(struct ccw_device *cdev,
struct chsc_pnso_area *pnso_area, u8 oc,
struct chsc_pnso_resume_token resume_token, int cnc);
+int ccw_device_get_cssid(struct ccw_device *cdev, u8 *cssid);
+int ccw_device_get_iid(struct ccw_device *cdev, u8 *iid);
+int ccw_device_get_chpid(struct ccw_device *cdev, int chp_idx, u8 *chpid);
+int ccw_device_get_chid(struct ccw_device *cdev, int chp_idx, u16 *chid);
#endif /* _S390_CCWDEV_H_ */
diff --git a/drivers/s390/cio/chsc.c b/drivers/s390/cio/chsc.c
index 8f764a295a51..38017c4a31e9 100644
--- a/drivers/s390/cio/chsc.c
+++ b/drivers/s390/cio/chsc.c
@@ -1116,7 +1116,7 @@ int chsc_enable_facility(int operation_code)
return ret;
}
-int __init chsc_get_cssid(int idx)
+int __init chsc_get_cssid_iid(int idx, u8 *cssid, u8 *iid)
{
struct {
struct chsc_header request;
@@ -1127,7 +1127,8 @@ int __init chsc_get_cssid(int idx)
u32 reserved2[3];
struct {
u8 cssid;
- u32 : 24;
+ u8 iid;
+ u32 : 16;
} list[0];
} *sdcal_area;
int ret;
@@ -1153,8 +1154,10 @@ int __init chsc_get_cssid(int idx)
}
if ((addr_t) &sdcal_area->list[idx] <
- (addr_t) &sdcal_area->response + sdcal_area->response.length)
- ret = sdcal_area->list[idx].cssid;
+ (addr_t) &sdcal_area->response + sdcal_area->response.length) {
+ *cssid = sdcal_area->list[idx].cssid;
+ *iid = sdcal_area->list[idx].iid;
+ }
else
ret = -ENODEV;
exit:
diff --git a/drivers/s390/cio/chsc.h b/drivers/s390/cio/chsc.h
index 7416957ba9f4..c2b83b68bc57 100644
--- a/drivers/s390/cio/chsc.h
+++ b/drivers/s390/cio/chsc.h
@@ -208,7 +208,7 @@ int chsc_scm_info(struct chsc_scm_info *scm_area, u64 token);
int chsc_pnso(struct subchannel_id schid, struct chsc_pnso_area *pnso_area,
u8 oc, struct chsc_pnso_resume_token resume_token, int cnc);
-int __init chsc_get_cssid(int idx);
+int __init chsc_get_cssid_iid(int idx, u8 *cssid, u8 *iid);
#ifdef CONFIG_SCM_BUS
int scm_update_information(void);
diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c
index aca022239b33..1981eb62d329 100644
--- a/drivers/s390/cio/css.c
+++ b/drivers/s390/cio/css.c
@@ -854,7 +854,7 @@ css_generate_pgid(struct channel_subsystem *css, u32 tod_high)
if (css_general_characteristics.mcss) {
css->global_pgid.pgid_high.ext_cssid.version = 0x80;
css->global_pgid.pgid_high.ext_cssid.cssid =
- (css->cssid < 0) ? 0 : css->cssid;
+ css->id_valid ? css->cssid : 0;
} else {
css->global_pgid.pgid_high.cpu_addr = stap();
}
@@ -877,7 +877,7 @@ static ssize_t real_cssid_show(struct device *dev, struct device_attribute *a,
{
struct channel_subsystem *css = to_css(dev);
- if (css->cssid < 0)
+ if (!css->id_valid)
return -EINVAL;
return sprintf(buf, "%x\n", css->cssid);
@@ -975,7 +975,12 @@ static int __init setup_css(int nr)
css->device.dma_mask = &css->device.coherent_dma_mask;
mutex_init(&css->mutex);
- css->cssid = chsc_get_cssid(nr);
+ ret = chsc_get_cssid_iid(nr, &css->cssid, &css->iid);
+ if (!ret) {
+ css->id_valid = true;
+ pr_info("Partition identifier %01x.%01x\n", css->cssid,
+ css->iid);
+ }
css_generate_pgid(css, (u32) (get_tod_clock() >> 32));
ret = device_register(&css->device);
diff --git a/drivers/s390/cio/css.h b/drivers/s390/cio/css.h
index 8d832900a63d..3f322ea0f498 100644
--- a/drivers/s390/cio/css.h
+++ b/drivers/s390/cio/css.h
@@ -115,7 +115,9 @@ extern int for_each_subchannel(int(*fn)(struct subchannel_id, void *), void *);
void css_update_ssd_info(struct subchannel *sch);
struct channel_subsystem {
- int cssid;
+ u8 cssid;
+ u8 iid;
+ bool id_valid; /* cssid,iid */
struct channel_path *chps[__MAX_CHPID + 1];
struct device device;
struct pgid global_pgid;
diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c
index cdf44f398957..0fe7b2f2e7f5 100644
--- a/drivers/s390/cio/device_ops.c
+++ b/drivers/s390/cio/device_ops.c
@@ -733,6 +733,91 @@ int ccw_device_pnso(struct ccw_device *cdev,
}
EXPORT_SYMBOL_GPL(ccw_device_pnso);
+/**
+ * ccw_device_get_cssid() - obtain Channel Subsystem ID
+ * @cdev: device to obtain the CSSID for
+ * @cssid: The resulting Channel Subsystem ID
+ */
+int ccw_device_get_cssid(struct ccw_device *cdev, u8 *cssid)
+{
+ struct device *sch_dev = cdev->dev.parent;
+ struct channel_subsystem *css = to_css(sch_dev->parent);
+
+ if (css->id_valid)
+ *cssid = css->cssid;
+ return css->id_valid ? 0 : -ENODEV;
+}
+EXPORT_SYMBOL_GPL(ccw_device_get_cssid);
+
+/**
+ * ccw_device_get_iid() - obtain MIF-image ID
+ * @cdev: device to obtain the MIF-image ID for
+ * @iid: The resulting MIF-image ID
+ */
+int ccw_device_get_iid(struct ccw_device *cdev, u8 *iid)
+{
+ struct device *sch_dev = cdev->dev.parent;
+ struct channel_subsystem *css = to_css(sch_dev->parent);
+
+ if (css->id_valid)
+ *iid = css->iid;
+ return css->id_valid ? 0 : -ENODEV;
+}
+EXPORT_SYMBOL_GPL(ccw_device_get_iid);
+
+/**
+ * ccw_device_get_chpid() - obtain Channel Path ID
+ * @cdev: device to obtain the Channel Path ID for
+ * @chp_idx: Index of the channel path
+ * @chpid: The resulting Channel Path ID
+ */
+int ccw_device_get_chpid(struct ccw_device *cdev, int chp_idx, u8 *chpid)
+{
+ struct subchannel *sch = to_subchannel(cdev->dev.parent);
+ int mask;
+
+ if ((chp_idx < 0) || (chp_idx > 7))
+ return -EINVAL;
+ mask = 0x80 >> chp_idx;
+ if (!(sch->schib.pmcw.pim & mask))
+ return -ENODEV;
+
+ *chpid = sch->schib.pmcw.chpid[chp_idx];
+ return 0;
+}
+EXPORT_SYMBOL_GPL(ccw_device_get_chpid);
+
+/**
+ * ccw_device_get_chid() - obtain Channel ID associated with specified CHPID
+ * @cdev: device to obtain the Channel ID for
+ * @chp_idx: Index of the channel path
+ * @chid: The resulting Channel ID
+ */
+int ccw_device_get_chid(struct ccw_device *cdev, int chp_idx, u16 *chid)
+{
+ struct chp_id cssid_chpid;
+ struct channel_path *chp;
+ int rc;
+
+ chp_id_init(&cssid_chpid);
+ rc = ccw_device_get_chpid(cdev, chp_idx, &cssid_chpid.id);
+ if (rc)
+ return rc;
+ chp = chpid_to_chp(cssid_chpid);
+ if (!chp)
+ return -ENODEV;
+
+ mutex_lock(&chp->lock);
+ if (chp->desc_fmt1.flags & 0x10)
+ *chid = chp->desc_fmt1.chid;
+ else
+ rc = -ENODEV;
+ mutex_unlock(&chp->lock);
+
+ return rc;
+}
+EXPORT_SYMBOL_GPL(ccw_device_get_chid);
+
/*
* Allocate zeroed dma coherent 31 bit addressable memory using
* the subchannels dma pool. Maximal size of allocation supported
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH net-next 3/8] s390/qeth: Detect PNSO OC3 capability
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 1/8] s390/cio: Add new Operation Code OC3 to PNSO Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 2/8] s390/cio: Helper functions to read CSSID, IID, and CHID Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 4/8] s390/qeth: Translate address events into switchdev notifiers Julian Wiedmann
` (5 subsequent siblings)
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
This patch detects whether device-to-bridge-notification, provided
by the Perform Network Subchannel Operation (PNSO) operation code
ADDR_INFO (OC3), is supported by this card. A following patch will
map this to the learning_sync bridgeport flag, so we store it in
priv->brport_hw_features in bridgeport flag format.
Only IQD cards provide PNSO.
There is a feature bit to indicate whether the machine provides OC3,
unfortunately it is not set on old machines.
So PNSO is called to find out. As this will disable notification
and is exclusive with bridgeport_notification, this must be done
during card initialisation before previous settings are restored.
PNSO functionality requires some configuration values that are added to
the qeth_card.info structure. Some helper functions are defined to fill
them out when the card is brought online and some other places are
adapted, that can also benefit from these fields.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
---
drivers/s390/net/qeth_core.h | 10 ++++++-
drivers/s390/net/qeth_core_main.c | 40 ++++++++++++++++++++-------
drivers/s390/net/qeth_l2_main.c | 45 +++++++++++++++++++++++++++++++
3 files changed, 85 insertions(+), 10 deletions(-)
diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h
index da46af682af8..14c583b5ea11 100644
--- a/drivers/s390/net/qeth_core.h
+++ b/drivers/s390/net/qeth_core.h
@@ -684,9 +684,16 @@ enum qeth_pnso_mode {
struct qeth_card_info {
unsigned short unit_addr2;
unsigned short cula;
- u8 chpid;
__u16 func_level;
char mcl_level[QETH_MCL_LENGTH + 1];
+ /* doubleword below corresponds to net_if_token */
+ u16 ddev_devno;
+ u8 cssid;
+ u8 iid;
+ u8 ssid;
+ u8 chpid;
+ u16 chid;
+ u8 ids_valid:1; /* cssid,iid,chid */
u8 dev_addr_is_registered:1;
u8 open_when_online:1;
u8 promisc_mode:1;
@@ -780,6 +787,7 @@ struct qeth_switch_info {
struct qeth_priv {
unsigned int rx_copybreak;
+ u32 brport_hw_features;
};
#define QETH_NAPI_WEIGHT NAPI_POLL_WEIGHT
diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c
index e19640bc6daa..7cd0cbf8a4f0 100644
--- a/drivers/s390/net/qeth_core_main.c
+++ b/drivers/s390/net/qeth_core_main.c
@@ -2311,12 +2311,10 @@ static void qeth_idx_setup_activate_cmd(struct qeth_card *card,
u16 addr = (card->info.cula << 8) + card->info.unit_addr2;
u8 port = ((u8)card->dev->dev_port) | 0x80;
struct ccw1 *ccw = __ccw_from_cmd(iob);
- struct ccw_dev_id dev_id;
qeth_setup_ccw(&ccw[0], CCW_CMD_WRITE, CCW_FLAG_CC, IDX_ACTIVATE_SIZE,
iob->data);
qeth_setup_ccw(&ccw[1], CCW_CMD_READ, 0, iob->length, iob->data);
- ccw_device_get_id(CARD_DDEV(card), &dev_id);
iob->finalize = qeth_idx_finalize_cmd;
port |= QETH_IDX_ACT_INVAL_FRAME;
@@ -2325,7 +2323,7 @@ static void qeth_idx_setup_activate_cmd(struct qeth_card *card,
&card->token.issuer_rm_w, QETH_MPC_TOKEN_LENGTH);
memcpy(QETH_IDX_ACT_FUNC_LEVEL(iob->data),
&card->info.func_level, 2);
- memcpy(QETH_IDX_ACT_QDIO_DEV_CUA(iob->data), &dev_id.devno, 2);
+ memcpy(QETH_IDX_ACT_QDIO_DEV_CUA(iob->data), &card->info.ddev_devno, 2);
memcpy(QETH_IDX_ACT_QDIO_DEV_REALADDR(iob->data), &addr, 2);
}
@@ -2599,7 +2597,6 @@ static int qeth_ulp_setup(struct qeth_card *card)
{
__u16 temp;
struct qeth_cmd_buffer *iob;
- struct ccw_dev_id dev_id;
QETH_CARD_TEXT(card, 2, "ulpsetup");
@@ -2614,8 +2611,7 @@ static int qeth_ulp_setup(struct qeth_card *card)
memcpy(QETH_ULP_SETUP_FILTER_TOKEN(iob->data),
&card->token.ulp_filter_r, QETH_MPC_TOKEN_LENGTH);
- ccw_device_get_id(CARD_DDEV(card), &dev_id);
- memcpy(QETH_ULP_SETUP_CUA(iob->data), &dev_id.devno, 2);
+ memcpy(QETH_ULP_SETUP_CUA(iob->data), &card->info.ddev_devno, 2);
temp = (card->info.cula << 8) + card->info.unit_addr2;
memcpy(QETH_ULP_SETUP_REAL_DEVADDR(iob->data), &temp, 2);
return qeth_send_control_data(card, iob, qeth_ulp_setup_cb, NULL);
@@ -4920,7 +4916,6 @@ int qeth_vm_request_mac(struct qeth_card *card)
{
struct diag26c_mac_resp *response;
struct diag26c_mac_req *request;
- struct ccw_dev_id id;
int rc;
QETH_CARD_TEXT(card, 2, "vmreqmac");
@@ -4932,11 +4927,10 @@ int qeth_vm_request_mac(struct qeth_card *card)
goto out;
}
- ccw_device_get_id(CARD_DDEV(card), &id);
request->resp_buf_len = sizeof(*response);
request->resp_version = DIAG26C_VERSION2;
request->op_code = DIAG26C_GET_MAC;
- request->devno = id.devno;
+ request->devno = card->info.ddev_devno;
QETH_DBF_HEX(CTRL, 2, request, sizeof(*request));
rc = diag26c(request, response, DIAG26C_MAC_SERVICES);
@@ -5017,6 +5011,33 @@ static void qeth_determine_capabilities(struct qeth_card *card)
return;
}
+static void qeth_read_ccw_conf_data(struct qeth_card *card)
+{
+ struct qeth_card_info *info = &card->info;
+ struct ccw_device *cdev = CARD_DDEV(card);
+ struct ccw_dev_id dev_id;
+
+ QETH_CARD_TEXT(card, 2, "ccwconfd");
+ ccw_device_get_id(cdev, &dev_id);
+
+ info->ddev_devno = dev_id.devno;
+ info->ids_valid = !ccw_device_get_cssid(cdev, &info->cssid) &&
+ !ccw_device_get_iid(cdev, &info->iid) &&
+ !ccw_device_get_chid(cdev, 0, &info->chid);
+ info->ssid = dev_id.ssid;
+
+ dev_info(&card->gdev->dev, "CHID: %x CHPID: %x\n",
+ info->chid, info->chpid);
+
+ QETH_CARD_TEXT_(card, 3, "devn%x", info->ddev_devno);
+ QETH_CARD_TEXT_(card, 3, "cssid:%x", info->cssid);
+ QETH_CARD_TEXT_(card, 3, "iid:%x", info->iid);
+ QETH_CARD_TEXT_(card, 3, "ssid:%x", info->ssid);
+ QETH_CARD_TEXT_(card, 3, "chpid:%x", info->chpid);
+ QETH_CARD_TEXT_(card, 3, "chid:%x", info->chid);
+ QETH_CARD_TEXT_(card, 3, "idval%x", info->ids_valid);
+}
+
static int qeth_qdio_establish(struct qeth_card *card)
{
struct qdio_buffer **out_sbal_ptrs[QETH_MAX_OUT_QUEUES];
@@ -5185,6 +5206,7 @@ int qeth_core_hardsetup_card(struct qeth_card *card, bool *carrier_ok)
}
qeth_determine_capabilities(card);
+ qeth_read_ccw_conf_data(card);
qeth_idx_init(card);
rc = qeth_idx_activate_read_channel(card);
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 2ab130d5c42d..7cba3d0035bf 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -17,10 +17,12 @@
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/etherdevice.h>
+#include <linux/if_bridge.h>
#include <linux/list.h>
#include <linux/hash.h>
#include <linux/hashtable.h>
#include <asm/chsc.h>
+#include <asm/css_chars.h>
#include <asm/setup.h>
#include "qeth_core.h"
#include "qeth_l2.h"
@@ -826,6 +828,46 @@ static void qeth_l2_setup_bridgeport_attrs(struct qeth_card *card)
}
}
+/**
+ * qeth_l2_detect_dev2br_support() -
+ * Detect whether this card supports 'dev to bridge fdb network address
+ * change notification' and thus can support the learning_sync bridgeport
+ * attribute
+ * @card: qeth_card structure pointer
+ *
+ * This is a destructive test and must be called before dev2br or
+ * bridgeport address notification is enabled!
+ */
+static void qeth_l2_detect_dev2br_support(struct qeth_card *card)
+{
+ struct qeth_priv *priv = netdev_priv(card->dev);
+ bool dev2br_supported;
+ int rc;
+
+ QETH_CARD_TEXT(card, 2, "d2brsup");
+ if (!IS_IQD(card))
+ return;
+
+ /* dev2br requires valid cssid,iid,chid */
+ if (!card->info.ids_valid) {
+ dev2br_supported = false;
+ } else if (css_general_characteristics.enarf) {
+ dev2br_supported = true;
+ } else {
+ /* Old machines don't have the feature bit:
+ * Probe by testing whether a disable succeeds
+ */
+ rc = qeth_l2_pnso(card, PNSO_OC_NET_ADDR_INFO, 0, NULL, NULL);
+ dev2br_supported = !rc;
+ }
+ QETH_CARD_TEXT_(card, 2, "D2Bsup%02x", dev2br_supported);
+
+ if (dev2br_supported)
+ priv->brport_hw_features |= BR_LEARNING_SYNC;
+ else
+ priv->brport_hw_features &= ~BR_LEARNING_SYNC;
+}
+
static int qeth_l2_set_online(struct qeth_card *card)
{
struct ccwgroup_device *gdev = card->gdev;
@@ -840,6 +882,9 @@ static int qeth_l2_set_online(struct qeth_card *card)
goto out_remove;
}
+ /* query before bridgeport_notification may be enabled */
+ qeth_l2_detect_dev2br_support(card);
+
mutex_lock(&card->sbp_lock);
qeth_bridgeport_query_support(card);
if (card->options.sbp.supported_funcs) {
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH net-next 4/8] s390/qeth: Translate address events into switchdev notifiers
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
` (2 preceding siblings ...)
2020-09-10 17:23 ` [PATCH net-next 3/8] s390/qeth: Detect PNSO OC3 capability Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier Julian Wiedmann
` (4 subsequent siblings)
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
A qeth-l2 HiperSockets card can show switch-ish behaviour in the sense,
that it can report all MACs that are reachable via this interface. Just
like a switch device, it can notify the software bridge about changes
to its fdb. This patch exploits this device-to-bridge-notification and
extracts the relevant information from the hardware events to generate
notifications to an attached software bridge.
There are 2 sources for this information:
1) The reply message of Perform-Network-Subchannel-Operations (PNSO)
(operation code ADDR_INFO) reports all addresses that are currently
reachable (implemented in a later patch).
2) As long as device-to-bridge-notification is enabled, hardware will
generate address change notification events, whenever the content of
the hardware fdb changes (this patch).
The bridge_hostnotify feature (PNSO operation code BRIDGE_INFO) uses
the same address change notification events. We need to distinguish
between qeth_pnso_mode QETH_PNSO_BRIDGEPORT and QETH_PNSO_ADDR_INFO
and call a different handler. In both cases deadlocks must be
prevented, if the workqueue is drained under lock and QETH_PNSO_NONE,
when notification is disabled.
bridge_hostnotify generates udev events, there is no intend to do the same
for dev2br. Instead this patch will generate SWITCHDEV_FDB_ADD_TO_BRIDGE
and SWITCHDEV_FDB_DEL_TO_BRIDGE notifications, that will cause the
software bridge to add (or delete) entries to its fdb as 'extern_learn
offload'.
Documentation/networking/switchdev.txt proposes to add
"depends NET_SWITCHDEV" to driver's Kconfig. This is not done here,
so even in absence of the NET_SWITCHDEV module, the QETH_L2 module will
still be built, but then the switchdev notifiers will have no effect.
No VLAN filtering is done on the entries and VLAN information is not
passed on to the bridge fdb entries. This could be added later.
For now VLAN interfaces can be defined on the upper bridge interface.
Multicast entries are not passed on to the bridge fdb.
This could be added later. For now mcast flooding can be used in the
bridge.
The card reports all MACs that are in its FDB, but we must not pass on
MACs that are registered for this interface.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
---
drivers/s390/net/qeth_core.h | 1 +
drivers/s390/net/qeth_l2_main.c | 110 +++++++++++++++++++++++++++++++-
2 files changed, 110 insertions(+), 1 deletion(-)
diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h
index 14c583b5ea11..4c8134a953c9 100644
--- a/drivers/s390/net/qeth_core.h
+++ b/drivers/s390/net/qeth_core.h
@@ -677,6 +677,7 @@ struct qeth_card_blkt {
enum qeth_pnso_mode {
QETH_PNSO_NONE,
QETH_PNSO_BRIDGEPORT,
+ QETH_PNSO_ADDR_INFO,
};
#define QETH_BROADCAST_WITH_ECHO 0x01
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 7cba3d0035bf..fffbc50cadc6 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -21,6 +21,7 @@
#include <linux/list.h>
#include <linux/hash.h>
#include <linux/hashtable.h>
+#include <net/switchdev.h>
#include <asm/chsc.h>
#include <asm/css_chars.h>
#include <asm/setup.h>
@@ -709,6 +710,68 @@ static int qeth_l2_pnso(struct qeth_card *card, u8 oc, int cnc,
return rc;
}
+static bool qeth_is_my_net_if_token(struct qeth_card *card,
+ struct net_if_token *token)
+{
+ return ((card->info.ddev_devno == token->devnum) &&
+ (card->info.cssid == token->cssid) &&
+ (card->info.iid == token->iid) &&
+ (card->info.ssid == token->ssid) &&
+ (card->info.chpid == token->chpid) &&
+ (card->info.chid == token->chid));
+}
+
+/**
+ * qeth_l2_dev2br_fdb_notify() - update fdb of master bridge
+ * @card: qeth_card structure pointer
+ * @code: event bitmask: high order bit 0x80 set to
+ * 1 - removal of an object
+ * 0 - addition of an object
+ * Object type(s):
+ * 0x01 - VLAN, 0x02 - MAC, 0x03 - VLAN and MAC
+ * @token: "network token" structure identifying 'physical' location
+ * of the target
+ * @addr_lnid: structure with MAC address and VLAN ID of the target
+ */
+static void qeth_l2_dev2br_fdb_notify(struct qeth_card *card, u8 code,
+ struct net_if_token *token,
+ struct mac_addr_lnid *addr_lnid)
+{
+ struct switchdev_notifier_fdb_info info;
+ u8 ntfy_mac[ETH_ALEN];
+
+ ether_addr_copy(ntfy_mac, addr_lnid->mac);
+ /* Ignore VLAN only changes */
+ if (!(code & IPA_ADDR_CHANGE_CODE_MACADDR))
+ return;
+ /* Ignore mcast entries */
+ if (is_multicast_ether_addr(ntfy_mac))
+ return;
+ /* Ignore my own addresses */
+ if (qeth_is_my_net_if_token(card, token))
+ return;
+
+ info.addr = ntfy_mac;
+ /* don't report VLAN IDs */
+ info.vid = 0;
+ info.added_by_user = false;
+ info.offloaded = true;
+
+ if (code & IPA_ADDR_CHANGE_CODE_REMOVAL) {
+ call_switchdev_notifiers(SWITCHDEV_FDB_DEL_TO_BRIDGE,
+ card->dev, &info.info, NULL);
+ QETH_CARD_TEXT(card, 4, "andelmac");
+ QETH_CARD_TEXT_(card, 4,
+ "mc%012lx", ether_addr_to_u64(ntfy_mac));
+ } else {
+ call_switchdev_notifiers(SWITCHDEV_FDB_ADD_TO_BRIDGE,
+ card->dev, &info.info, NULL);
+ QETH_CARD_TEXT(card, 4, "anaddmac");
+ QETH_CARD_TEXT_(card, 4,
+ "mc%012lx", ether_addr_to_u64(ntfy_mac));
+ }
+}
+
static const struct net_device_ops qeth_l2_netdev_ops = {
.ndo_open = qeth_open,
.ndo_stop = qeth_stop,
@@ -1216,6 +1279,48 @@ struct qeth_addr_change_data {
struct qeth_ipacmd_addr_change ac_event;
};
+static void qeth_l2_dev2br_worker(struct work_struct *work)
+{
+ struct delayed_work *dwork = to_delayed_work(work);
+ struct qeth_addr_change_data *data;
+ struct qeth_card *card;
+ unsigned int i;
+
+ data = container_of(dwork, struct qeth_addr_change_data, dwork);
+ card = data->card;
+
+ QETH_CARD_TEXT(card, 4, "dev2brew");
+
+ if (READ_ONCE(card->info.pnso_mode) == QETH_PNSO_NONE)
+ goto free;
+
+ /* Potential re-config in progress, try again later: */
+ if (!rtnl_trylock()) {
+ queue_delayed_work(card->event_wq, dwork,
+ msecs_to_jiffies(100));
+ return;
+ }
+
+ if (data->ac_event.lost_event_mask) {
+ QETH_DBF_MESSAGE(3,
+ "Address change notification overflow on device %x\n",
+ CARD_DEVID(card));
+ } else {
+ for (i = 0; i < data->ac_event.num_entries; i++) {
+ struct qeth_ipacmd_addr_change_entry *entry =
+ &data->ac_event.entry[i];
+ qeth_l2_dev2br_fdb_notify(card,
+ entry->change_code,
+ &entry->token,
+ &entry->addr_lnid);
+ }
+ }
+ rtnl_unlock();
+
+free:
+ kfree(data);
+}
+
static void qeth_addr_change_event_worker(struct work_struct *work)
{
struct delayed_work *dwork = to_delayed_work(work);
@@ -1298,7 +1403,10 @@ static void qeth_addr_change_event(struct qeth_card *card,
QETH_CARD_TEXT(card, 2, "ACNalloc");
return;
}
- INIT_DELAYED_WORK(&data->dwork, qeth_addr_change_event_worker);
+ if (card->info.pnso_mode == QETH_PNSO_BRIDGEPORT)
+ INIT_DELAYED_WORK(&data->dwork, qeth_addr_change_event_worker);
+ else
+ INIT_DELAYED_WORK(&data->dwork, qeth_l2_dev2br_worker);
data->card = card;
memcpy(&data->ac_event, hostevs,
sizeof(struct qeth_ipacmd_addr_change) + extrasize);
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
` (3 preceding siblings ...)
2020-09-10 17:23 ` [PATCH net-next 4/8] s390/qeth: Translate address events into switchdev notifiers Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-14 21:38 ` David Miller
` (2 more replies)
2020-09-10 17:23 ` [PATCH net-next 6/8] s390/qeth: Reset address notification in case of buffer overflow Julian Wiedmann
` (3 subsequent siblings)
8 siblings, 3 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
so the switchdev can notifiy the bridge to flush non-permanent fdb entries
for this port. This is useful whenever the hardware fdb of the switchdev
is reset, but the netdev and the bridgeport are not deleted.
Note that this has the same effect as the IFLA_BRPORT_FLUSH attribute.
CC: Jiri Pirko <jiri@resnulli.us>
CC: Ivan Vecera <ivecera@redhat.com>
CC: Roopa Prabhu <roopa@nvidia.com>
CC: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
---
include/net/switchdev.h | 1 +
net/bridge/br.c | 5 +++++
2 files changed, 6 insertions(+)
diff --git a/include/net/switchdev.h b/include/net/switchdev.h
index ff2246914301..53e8b4994296 100644
--- a/include/net/switchdev.h
+++ b/include/net/switchdev.h
@@ -203,6 +203,7 @@ enum switchdev_notifier_type {
SWITCHDEV_FDB_ADD_TO_DEVICE,
SWITCHDEV_FDB_DEL_TO_DEVICE,
SWITCHDEV_FDB_OFFLOADED,
+ SWITCHDEV_FDB_FLUSH_TO_BRIDGE,
SWITCHDEV_PORT_OBJ_ADD, /* Blocking. */
SWITCHDEV_PORT_OBJ_DEL, /* Blocking. */
diff --git a/net/bridge/br.c b/net/bridge/br.c
index b6fe30e3768f..401eeb9142eb 100644
--- a/net/bridge/br.c
+++ b/net/bridge/br.c
@@ -183,6 +183,11 @@ static int br_switchdev_event(struct notifier_block *unused,
br_fdb_offloaded_set(br, p, fdb_info->addr,
fdb_info->vid, fdb_info->offloaded);
break;
+ case SWITCHDEV_FDB_FLUSH_TO_BRIDGE:
+ fdb_info = ptr;
+ /* Don't delete static entries */
+ br_fdb_delete_by_port(br, p, fdb_info->vid, 0);
+ break;
}
out:
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* Re: [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier
2020-09-10 17:23 ` [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier Julian Wiedmann
@ 2020-09-14 21:38 ` David Miller
2020-09-15 7:20 ` Nikolay Aleksandrov
2020-09-15 8:19 ` Ivan Vecera
2 siblings, 0 replies; 13+ messages in thread
From: David Miller @ 2020-09-14 21:38 UTC (permalink / raw)
To: jwi
Cc: kuba, netdev, linux-s390, hca, ubraun, kgraul, wintera, roopa,
nikolay, jiri, ivecera
From: Julian Wiedmann <jwi@linux.ibm.com>
Date: Thu, 10 Sep 2020 19:23:48 +0200
> From: Alexandra Winter <wintera@linux.ibm.com>
>
> so the switchdev can notifiy the bridge to flush non-permanent fdb entries
> for this port. This is useful whenever the hardware fdb of the switchdev
> is reset, but the netdev and the bridgeport are not deleted.
>
> Note that this has the same effect as the IFLA_BRPORT_FLUSH attribute.
>
> CC: Jiri Pirko <jiri@resnulli.us>
> CC: Ivan Vecera <ivecera@redhat.com>
> CC: Roopa Prabhu <roopa@nvidia.com>
> CC: Nikolay Aleksandrov <nikolay@nvidia.com>
> Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
> Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
This still needs review by bridge experts.
Thank you.
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier
2020-09-10 17:23 ` [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier Julian Wiedmann
2020-09-14 21:38 ` David Miller
@ 2020-09-15 7:20 ` Nikolay Aleksandrov
2020-09-15 8:19 ` Ivan Vecera
2 siblings, 0 replies; 13+ messages in thread
From: Nikolay Aleksandrov @ 2020-09-15 7:20 UTC (permalink / raw)
To: davem@davemloft.net, jwi@linux.ibm.com, kuba@kernel.org
Cc: kgraul@linux.ibm.com, hca@linux.ibm.com,
linux-s390@vger.kernel.org, wintera@linux.ibm.com,
ubraun@linux.ibm.com, jiri@resnulli.us, netdev@vger.kernel.org,
Roopa Prabhu, ivecera@redhat.com
On Thu, 2020-09-10 at 19:23 +0200, Julian Wiedmann wrote:
> From: Alexandra Winter <wintera@linux.ibm.com>
>
> so the switchdev can notifiy the bridge to flush non-permanent fdb entries
> for this port. This is useful whenever the hardware fdb of the switchdev
> is reset, but the netdev and the bridgeport are not deleted.
>
> Note that this has the same effect as the IFLA_BRPORT_FLUSH attribute.
>
> CC: Jiri Pirko <jiri@resnulli.us>
> CC: Ivan Vecera <ivecera@redhat.com>
> CC: Roopa Prabhu <roopa@nvidia.com>
> CC: Nikolay Aleksandrov <nikolay@nvidia.com>
> Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
> Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
> ---
> include/net/switchdev.h | 1 +
> net/bridge/br.c | 5 +++++
> 2 files changed, 6 insertions(+)
>
> diff --git a/include/net/switchdev.h b/include/net/switchdev.h
> index ff2246914301..53e8b4994296 100644
> --- a/include/net/switchdev.h
> +++ b/include/net/switchdev.h
> @@ -203,6 +203,7 @@ enum switchdev_notifier_type {
> SWITCHDEV_FDB_ADD_TO_DEVICE,
> SWITCHDEV_FDB_DEL_TO_DEVICE,
> SWITCHDEV_FDB_OFFLOADED,
> + SWITCHDEV_FDB_FLUSH_TO_BRIDGE,
>
> SWITCHDEV_PORT_OBJ_ADD, /* Blocking. */
> SWITCHDEV_PORT_OBJ_DEL, /* Blocking. */
> diff --git a/net/bridge/br.c b/net/bridge/br.c
> index b6fe30e3768f..401eeb9142eb 100644
> --- a/net/bridge/br.c
> +++ b/net/bridge/br.c
> @@ -183,6 +183,11 @@ static int br_switchdev_event(struct notifier_block *unused,
> br_fdb_offloaded_set(br, p, fdb_info->addr,
> fdb_info->vid, fdb_info->offloaded);
> break;
> + case SWITCHDEV_FDB_FLUSH_TO_BRIDGE:
> + fdb_info = ptr;
> + /* Don't delete static entries */
> + br_fdb_delete_by_port(br, p, fdb_info->vid, 0);
> + break;
> }
>
> out:
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
^ permalink raw reply [flat|nested] 13+ messages in thread* Re: [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier
2020-09-10 17:23 ` [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier Julian Wiedmann
2020-09-14 21:38 ` David Miller
2020-09-15 7:20 ` Nikolay Aleksandrov
@ 2020-09-15 8:19 ` Ivan Vecera
2 siblings, 0 replies; 13+ messages in thread
From: Ivan Vecera @ 2020-09-15 8:19 UTC (permalink / raw)
To: Julian Wiedmann
Cc: David Miller, Jakub Kicinski, netdev, linux-s390, Heiko Carstens,
Ursula Braun, Karsten Graul, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko
On Thu, 10 Sep 2020 19:23:48 +0200
Julian Wiedmann <jwi@linux.ibm.com> wrote:
> From: Alexandra Winter <wintera@linux.ibm.com>
>
> so the switchdev can notifiy the bridge to flush non-permanent fdb entries
> for this port. This is useful whenever the hardware fdb of the switchdev
> is reset, but the netdev and the bridgeport are not deleted.
>
> Note that this has the same effect as the IFLA_BRPORT_FLUSH attribute.
>
> CC: Jiri Pirko <jiri@resnulli.us>
> CC: Ivan Vecera <ivecera@redhat.com>
> CC: Roopa Prabhu <roopa@nvidia.com>
> CC: Nikolay Aleksandrov <nikolay@nvidia.com>
> Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
> Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
> ---
> include/net/switchdev.h | 1 +
> net/bridge/br.c | 5 +++++
> 2 files changed, 6 insertions(+)
>
> diff --git a/include/net/switchdev.h b/include/net/switchdev.h
> index ff2246914301..53e8b4994296 100644
> --- a/include/net/switchdev.h
> +++ b/include/net/switchdev.h
> @@ -203,6 +203,7 @@ enum switchdev_notifier_type {
> SWITCHDEV_FDB_ADD_TO_DEVICE,
> SWITCHDEV_FDB_DEL_TO_DEVICE,
> SWITCHDEV_FDB_OFFLOADED,
> + SWITCHDEV_FDB_FLUSH_TO_BRIDGE,
>
> SWITCHDEV_PORT_OBJ_ADD, /* Blocking. */
> SWITCHDEV_PORT_OBJ_DEL, /* Blocking. */
> diff --git a/net/bridge/br.c b/net/bridge/br.c
> index b6fe30e3768f..401eeb9142eb 100644
> --- a/net/bridge/br.c
> +++ b/net/bridge/br.c
> @@ -183,6 +183,11 @@ static int br_switchdev_event(struct notifier_block *unused,
> br_fdb_offloaded_set(br, p, fdb_info->addr,
> fdb_info->vid, fdb_info->offloaded);
> break;
> + case SWITCHDEV_FDB_FLUSH_TO_BRIDGE:
> + fdb_info = ptr;
> + /* Don't delete static entries */
> + br_fdb_delete_by_port(br, p, fdb_info->vid, 0);
> + break;
> }
>
> out:
Acked-by: Ivan Vecera <ivecera@redhat.com>
^ permalink raw reply [flat|nested] 13+ messages in thread
* [PATCH net-next 6/8] s390/qeth: Reset address notification in case of buffer overflow
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
` (4 preceding siblings ...)
2020-09-10 17:23 ` [PATCH net-next 5/8] bridge: Add SWITCHDEV_FDB_FLUSH_TO_BRIDGE notifier Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 7/8] s390/qeth: implement ndo_bridge_getlink for learning_sync Julian Wiedmann
` (2 subsequent siblings)
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
In case hardware sends more device-to-bridge-address-change notfications
than the qeth-l2 driver can handle, the hardware will send an overflow
event and then stop sending any events. It expects software to flush its
FDB and start over again. Re-enabling address-change-notification will
report all current addresses.
In order to re-enable address-change-notification this patch defines
the functions qeth_l2_dev2br_an_set() and qeth_l2_dev2br_an_set_cb
to enable or disable dev-to-bridge-address-notification.
A following patch will use the learning_sync bridgeport flag to trigger
enabling or disabling of address-change-notification, so we define
priv->brport_features to store the current setting. BRIDGE_INFO and
ADDR_INFO functionality are mutually exclusive, whereas ADDR_INFO and
qeth_l2_vnicc* can be used together.
Alternative implementations to handle buffer overflow:
Just re-enabling notification and adding all newly reported addresses
would cover any lost 'add' events, but not the lost 'delete' events.
Then these invalid addresses would stay in the bridge FDB as long as the
device exists.
Setting the net device down and up, would be an alternative, but is a bit
drastic. If the net device has many secondary addresses this will create
many delete/add events at its peers which could de-stabilize the
network segment.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
---
drivers/s390/net/qeth_core.h | 1 +
drivers/s390/net/qeth_l2.h | 2 +-
drivers/s390/net/qeth_l2_main.c | 114 +++++++++++++++++++++++++++++++-
drivers/s390/net/qeth_l2_sys.c | 16 ++---
4 files changed, 123 insertions(+), 10 deletions(-)
diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h
index 4c8134a953c9..2c14012ca35d 100644
--- a/drivers/s390/net/qeth_core.h
+++ b/drivers/s390/net/qeth_core.h
@@ -789,6 +789,7 @@ struct qeth_switch_info {
struct qeth_priv {
unsigned int rx_copybreak;
u32 brport_hw_features;
+ u32 brport_features;
};
#define QETH_NAPI_WEIGHT NAPI_POLL_WEIGHT
diff --git a/drivers/s390/net/qeth_l2.h b/drivers/s390/net/qeth_l2.h
index adf25c9fd2b3..cc95675c8bc4 100644
--- a/drivers/s390/net/qeth_l2.h
+++ b/drivers/s390/net/qeth_l2.h
@@ -23,7 +23,7 @@ int qeth_l2_vnicc_set_state(struct qeth_card *card, u32 vnicc, bool state);
int qeth_l2_vnicc_get_state(struct qeth_card *card, u32 vnicc, bool *state);
int qeth_l2_vnicc_set_timeout(struct qeth_card *card, u32 timeout);
int qeth_l2_vnicc_get_timeout(struct qeth_card *card, u32 *timeout);
-bool qeth_l2_vnicc_is_in_use(struct qeth_card *card);
+bool qeth_bridgeport_allowed(struct qeth_card *card);
struct qeth_mac {
u8 mac_addr[ETH_ALEN];
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index fffbc50cadc6..ef2962e03546 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -287,6 +287,22 @@ static void qeth_l2_set_pnso_mode(struct qeth_card *card,
drain_workqueue(card->event_wq);
}
+static void qeth_l2_dev2br_fdb_flush(struct qeth_card *card)
+{
+ struct switchdev_notifier_fdb_info info;
+
+ QETH_CARD_TEXT(card, 2, "fdbflush");
+
+ info.addr = NULL;
+ /* flush all VLANs: */
+ info.vid = 0;
+ info.added_by_user = false;
+ info.offloaded = true;
+
+ call_switchdev_notifiers(SWITCHDEV_FDB_FLUSH_TO_BRIDGE,
+ card->dev, &info.info, NULL);
+}
+
static void qeth_l2_stop_card(struct qeth_card *card)
{
QETH_CARD_TEXT(card, 2, "stopcard");
@@ -772,6 +788,54 @@ static void qeth_l2_dev2br_fdb_notify(struct qeth_card *card, u8 code,
}
}
+static void qeth_l2_dev2br_an_set_cb(void *priv,
+ struct chsc_pnso_naid_l2 *entry)
+{
+ u8 code = IPA_ADDR_CHANGE_CODE_MACADDR;
+ struct qeth_card *card = priv;
+
+ if (entry->addr_lnid.lnid < VLAN_N_VID)
+ code |= IPA_ADDR_CHANGE_CODE_VLANID;
+ qeth_l2_dev2br_fdb_notify(card, code,
+ (struct net_if_token *)&entry->nit,
+ (struct mac_addr_lnid *)&entry->addr_lnid);
+}
+
+/**
+ * qeth_l2_dev2br_an_set() -
+ * Enable or disable 'dev to bridge network address notification'
+ * @card: qeth_card structure pointer
+ * @enable: Enable or disable 'dev to bridge network address notification'
+ *
+ * Returns negative errno-compatible error indication or 0 on success.
+ *
+ * On enable, emits a series of address notifications for all
+ * currently registered hosts.
+ *
+ * Must be called under rtnl_lock
+ */
+static int qeth_l2_dev2br_an_set(struct qeth_card *card, bool enable)
+{
+ int rc;
+
+ if (enable) {
+ QETH_CARD_TEXT(card, 2, "anseton");
+ rc = qeth_l2_pnso(card, PNSO_OC_NET_ADDR_INFO, 1,
+ qeth_l2_dev2br_an_set_cb, card);
+ if (rc == -EAGAIN)
+ /* address notification enabled, but inconsistent
+ * addresses reported -> disable address notification
+ */
+ qeth_l2_pnso(card, PNSO_OC_NET_ADDR_INFO, 0,
+ NULL, NULL);
+ } else {
+ QETH_CARD_TEXT(card, 2, "ansetoff");
+ rc = qeth_l2_pnso(card, PNSO_OC_NET_ADDR_INFO, 0, NULL, NULL);
+ }
+
+ return rc;
+}
+
static const struct net_device_ops qeth_l2_netdev_ops = {
.ndo_open = qeth_open,
.ndo_stop = qeth_stop,
@@ -1284,10 +1348,13 @@ static void qeth_l2_dev2br_worker(struct work_struct *work)
struct delayed_work *dwork = to_delayed_work(work);
struct qeth_addr_change_data *data;
struct qeth_card *card;
+ struct qeth_priv *priv;
unsigned int i;
+ int rc;
data = container_of(dwork, struct qeth_addr_change_data, dwork);
card = data->card;
+ priv = netdev_priv(card->dev);
QETH_CARD_TEXT(card, 4, "dev2brew");
@@ -1300,11 +1367,39 @@ static void qeth_l2_dev2br_worker(struct work_struct *work)
msecs_to_jiffies(100));
return;
}
+ if (!netif_device_present(card->dev))
+ goto out_unlock;
if (data->ac_event.lost_event_mask) {
QETH_DBF_MESSAGE(3,
"Address change notification overflow on device %x\n",
CARD_DEVID(card));
+ /* Card fdb and bridge fdb are out of sync, card has stopped
+ * notifications (no need to drain_workqueue). Purge all
+ * 'extern_learn' entries from the parent bridge and restart
+ * the notifications.
+ */
+ qeth_l2_dev2br_fdb_flush(card);
+ rc = qeth_l2_dev2br_an_set(card, true);
+ if (rc) {
+ /* TODO: if we want to retry after -EAGAIN, be
+ * aware there could be stale entries in the
+ * workqueue now, that need to be drained.
+ * For now we give up:
+ */
+ netdev_err(card->dev,
+ "bridge learning_sync failed to recover: %d\n",
+ rc);
+ WRITE_ONCE(card->info.pnso_mode,
+ QETH_PNSO_NONE);
+ /* To remove fdb entries reported by an_set: */
+ qeth_l2_dev2br_fdb_flush(card);
+ priv->brport_features ^= BR_LEARNING_SYNC;
+ } else {
+ QETH_DBF_MESSAGE(3,
+ "Address Notification resynced on device %x\n",
+ CARD_DEVID(card));
+ }
} else {
for (i = 0; i < data->ac_event.num_entries; i++) {
struct qeth_ipacmd_addr_change_entry *entry =
@@ -1315,6 +1410,8 @@ static void qeth_l2_dev2br_worker(struct work_struct *work)
&entry->addr_lnid);
}
}
+
+out_unlock:
rtnl_unlock();
free:
@@ -2035,7 +2132,7 @@ int qeth_l2_vnicc_get_timeout(struct qeth_card *card, u32 *timeout)
}
/* check if VNICC is currently enabled */
-bool qeth_l2_vnicc_is_in_use(struct qeth_card *card)
+static bool _qeth_l2_vnicc_is_in_use(struct qeth_card *card)
{
if (!card->options.vnicc.sup_chars)
return false;
@@ -2050,6 +2147,21 @@ bool qeth_l2_vnicc_is_in_use(struct qeth_card *card)
return true;
}
+/**
+ * qeth_bridgeport_allowed - are any qeth_bridgeport functions allowed?
+ * @card: qeth_card structure pointer
+ *
+ * qeth_bridgeport functionality is mutually exclusive with usage of the
+ * VNIC Characteristics and dev2br address notifications
+ */
+bool qeth_bridgeport_allowed(struct qeth_card *card)
+{
+ struct qeth_priv *priv = netdev_priv(card->dev);
+
+ return (!_qeth_l2_vnicc_is_in_use(card) &&
+ !(priv->brport_features & BR_LEARNING_SYNC));
+}
+
/* recover user timeout setting */
static bool qeth_l2_vnicc_recover_timeout(struct qeth_card *card, u32 vnicc,
u32 *timeout)
diff --git a/drivers/s390/net/qeth_l2_sys.c b/drivers/s390/net/qeth_l2_sys.c
index 4695d25e54f2..4ba3bc57263f 100644
--- a/drivers/s390/net/qeth_l2_sys.c
+++ b/drivers/s390/net/qeth_l2_sys.c
@@ -18,7 +18,7 @@ static ssize_t qeth_bridge_port_role_state_show(struct device *dev,
int rc = 0;
char *word;
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
return sprintf(buf, "n/a (VNIC characteristics)\n");
mutex_lock(&card->sbp_lock);
@@ -65,7 +65,7 @@ static ssize_t qeth_bridge_port_role_show(struct device *dev,
{
struct qeth_card *card = dev_get_drvdata(dev);
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
return sprintf(buf, "n/a (VNIC characteristics)\n");
return qeth_bridge_port_role_state_show(dev, attr, buf, 0);
@@ -90,7 +90,7 @@ static ssize_t qeth_bridge_port_role_store(struct device *dev,
mutex_lock(&card->conf_mutex);
mutex_lock(&card->sbp_lock);
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
rc = -EBUSY;
else if (card->options.sbp.reflect_promisc)
/* Forbid direct manipulation */
@@ -116,7 +116,7 @@ static ssize_t qeth_bridge_port_state_show(struct device *dev,
{
struct qeth_card *card = dev_get_drvdata(dev);
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
return sprintf(buf, "n/a (VNIC characteristics)\n");
return qeth_bridge_port_role_state_show(dev, attr, buf, 1);
@@ -131,7 +131,7 @@ static ssize_t qeth_bridgeport_hostnotification_show(struct device *dev,
struct qeth_card *card = dev_get_drvdata(dev);
int enabled;
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
return sprintf(buf, "n/a (VNIC characteristics)\n");
enabled = card->options.sbp.hostnotification;
@@ -153,7 +153,7 @@ static ssize_t qeth_bridgeport_hostnotification_store(struct device *dev,
mutex_lock(&card->conf_mutex);
mutex_lock(&card->sbp_lock);
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
rc = -EBUSY;
else if (qeth_card_hw_is_reachable(card)) {
rc = qeth_bridgeport_an_set(card, enable);
@@ -179,7 +179,7 @@ static ssize_t qeth_bridgeport_reflect_show(struct device *dev,
struct qeth_card *card = dev_get_drvdata(dev);
char *state;
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
return sprintf(buf, "n/a (VNIC characteristics)\n");
if (card->options.sbp.reflect_promisc) {
@@ -215,7 +215,7 @@ static ssize_t qeth_bridgeport_reflect_store(struct device *dev,
mutex_lock(&card->conf_mutex);
mutex_lock(&card->sbp_lock);
- if (qeth_l2_vnicc_is_in_use(card))
+ if (!qeth_bridgeport_allowed(card))
rc = -EBUSY;
else if (card->options.sbp.role != QETH_SBP_ROLE_NONE)
rc = -EPERM;
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH net-next 7/8] s390/qeth: implement ndo_bridge_getlink for learning_sync
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
` (5 preceding siblings ...)
2020-09-10 17:23 ` [PATCH net-next 6/8] s390/qeth: Reset address notification in case of buffer overflow Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-10 17:23 ` [PATCH net-next 8/8] s390/qeth: implement ndo_bridge_setlink " Julian Wiedmann
2020-09-15 20:22 ` [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 David Miller
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
Documentation/networking/switchdev.txt and 'man bridge' indicate that the
learning_sync bridge attribute is used to indicate whether a given
device will sync MAC addresses learned on its device port to a master
bridge FDB.
learning_sync attribute can not be read while interface is offline (down).
See
'commit e6e771b3d897 ("s390/qeth: detach netdevice while card is offline")'
We return EOPNOTSUPP and not EONODEV in this case, because EONOTSUPP is the
only rc that is tolerated by 'bridge -d link show'.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 23 ++++++++++++++++++++++-
1 file changed, 22 insertions(+), 1 deletion(-)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index ef2962e03546..338bc62556cf 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -33,6 +33,7 @@ static void qeth_bridge_state_change(struct qeth_card *card,
struct qeth_ipa_cmd *cmd);
static void qeth_addr_change_event(struct qeth_card *card,
struct qeth_ipa_cmd *cmd);
+static bool qeth_bridgeport_is_in_use(struct qeth_card *card);
static void qeth_l2_vnicc_set_defaults(struct qeth_card *card);
static void qeth_l2_vnicc_init(struct qeth_card *card);
static bool qeth_l2_vnicc_recover_timeout(struct qeth_card *card, u32 vnicc,
@@ -836,6 +837,25 @@ static int qeth_l2_dev2br_an_set(struct qeth_card *card, bool enable)
return rc;
}
+static int qeth_l2_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
+ struct net_device *dev, u32 filter_mask,
+ int nlflags)
+{
+ struct qeth_priv *priv = netdev_priv(dev);
+ struct qeth_card *card = dev->ml_priv;
+ u16 mode = BRIDGE_MODE_UNDEF;
+
+ /* Do not even show qeth devs that cannot do bridge_setlink */
+ if (!priv->brport_hw_features || !netif_device_present(dev) ||
+ qeth_bridgeport_is_in_use(card))
+ return -EOPNOTSUPP;
+
+ return ndo_dflt_bridge_getlink(skb, pid, seq, dev,
+ mode, priv->brport_features,
+ priv->brport_hw_features,
+ nlflags, filter_mask, NULL);
+}
+
static const struct net_device_ops qeth_l2_netdev_ops = {
.ndo_open = qeth_open,
.ndo_stop = qeth_stop,
@@ -851,7 +871,8 @@ static const struct net_device_ops qeth_l2_netdev_ops = {
.ndo_vlan_rx_kill_vid = qeth_l2_vlan_rx_kill_vid,
.ndo_tx_timeout = qeth_tx_timeout,
.ndo_fix_features = qeth_fix_features,
- .ndo_set_features = qeth_set_features
+ .ndo_set_features = qeth_set_features,
+ .ndo_bridge_getlink = qeth_l2_bridge_getlink,
};
static const struct net_device_ops qeth_osn_netdev_ops = {
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH net-next 8/8] s390/qeth: implement ndo_bridge_setlink for learning_sync
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
` (6 preceding siblings ...)
2020-09-10 17:23 ` [PATCH net-next 7/8] s390/qeth: implement ndo_bridge_getlink for learning_sync Julian Wiedmann
@ 2020-09-10 17:23 ` Julian Wiedmann
2020-09-15 20:22 ` [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 David Miller
8 siblings, 0 replies; 13+ messages in thread
From: Julian Wiedmann @ 2020-09-10 17:23 UTC (permalink / raw)
To: David Miller, Jakub Kicinski
Cc: netdev, linux-s390, Heiko Carstens, Ursula Braun, Karsten Graul,
Julian Wiedmann, Alexandra Winter, Roopa Prabhu,
Nikolay Aleksandrov, Jiri Pirko, Ivan Vecera
From: Alexandra Winter <wintera@linux.ibm.com>
Documentation/networking/switchdev.txt and 'man bridge' indicate that the
learning_sync bridge attribute is used to control whether a given
device will sync MAC addresses learned on its device port to a master
bridge FDB, where they will show up as 'extern_learn offload'. So we map
qeth_l2_dev2br_an_set() to the learning_sync bridge link attribute.
Turning off learning_sync will flush all extern_learn entries from the
bridge fdb and all pending events from the card's work queue.
When the hardware interface goes offline with learning_sync on
(e.g. for HW recovery), all extern_learn entries will be flushed from the
bridge fdb and all pending events from the card's work queue. When the
interface goes online again, it will send new notifications for all then
valid MACs. learning_sync attribute can not be modified while interface is
offline. See
'commit e6e771b3d897 ("s390/qeth: detach netdevice while card is offline")'
An alternative implementation would be to always offload the 'learning'
attribute of a software bridge to the hardware interface attached to it
and thus implicitly enable fdb notification. This was not chosen for 2
reasons:
1) In our case the software bridge is NOT a representation of a hardware
switch. It is just connected to a smart NIC that is able to inform
about the addresses attached to it. It is not necessarily using source
MAC learning for this and other bridgeports can be attached to other
NICs with different properties.
2) We want a means to enable this notification explicitly. There may be
cases where a bridgeport is set to 'learning', but we do not want to
enable the notification.
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 125 ++++++++++++++++++++++++++++++++
1 file changed, 125 insertions(+)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 338bc62556cf..54e02518ce08 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -306,6 +306,8 @@ static void qeth_l2_dev2br_fdb_flush(struct qeth_card *card)
static void qeth_l2_stop_card(struct qeth_card *card)
{
+ struct qeth_priv *priv = netdev_priv(card->dev);
+
QETH_CARD_TEXT(card, 2, "stopcard");
qeth_set_allowed_threads(card, 0, 1);
@@ -324,6 +326,12 @@ static void qeth_l2_stop_card(struct qeth_card *card)
qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
qeth_flush_local_addrs(card);
card->info.promisc_mode = 0;
+
+ if (priv->brport_features & BR_LEARNING_SYNC) {
+ rtnl_lock();
+ qeth_l2_dev2br_fdb_flush(card);
+ rtnl_unlock();
+ }
}
static int qeth_l2_request_initial_mac(struct qeth_card *card)
@@ -856,6 +864,89 @@ static int qeth_l2_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
nlflags, filter_mask, NULL);
}
+static const struct nla_policy qeth_brport_policy[IFLA_BRPORT_MAX + 1] = {
+ [IFLA_BRPORT_LEARNING_SYNC] = { .type = NLA_U8 },
+};
+
+/**
+ * qeth_l2_bridge_setlink() - set bridgeport attributes
+ * @dev: netdevice
+ * @nlh: netlink message header
+ * @flags: bridge flags (here: BRIDGE_FLAGS_SELF)
+ * @extack: extended ACK report struct
+ *
+ * Called under rtnl_lock
+ */
+static int qeth_l2_bridge_setlink(struct net_device *dev, struct nlmsghdr *nlh,
+ u16 flags, struct netlink_ext_ack *extack)
+{
+ struct qeth_priv *priv = netdev_priv(dev);
+ struct nlattr *bp_tb[IFLA_BRPORT_MAX + 1];
+ struct qeth_card *card = dev->ml_priv;
+ struct nlattr *attr, *nested_attr;
+ bool enable, has_protinfo = false;
+ int rem1, rem2;
+ int rc;
+
+ if (!netif_device_present(dev))
+ return -ENODEV;
+ if (!(priv->brport_hw_features))
+ return -EOPNOTSUPP;
+
+ nlmsg_for_each_attr(attr, nlh, sizeof(struct ifinfomsg), rem1) {
+ if (nla_type(attr) == IFLA_PROTINFO) {
+ rc = nla_parse_nested(bp_tb, IFLA_BRPORT_MAX, attr,
+ qeth_brport_policy, extack);
+ if (rc)
+ return rc;
+ has_protinfo = true;
+ } else if (nla_type(attr) == IFLA_AF_SPEC) {
+ nla_for_each_nested(nested_attr, attr, rem2) {
+ if (nla_type(nested_attr) == IFLA_BRIDGE_FLAGS)
+ continue;
+ NL_SET_ERR_MSG_ATTR(extack, nested_attr,
+ "Unsupported attribute");
+ return -EINVAL;
+ }
+ } else {
+ NL_SET_ERR_MSG_ATTR(extack, attr, "Unsupported attribute");
+ return -EINVAL;
+ }
+ }
+ if (!has_protinfo)
+ return 0;
+ if (!bp_tb[IFLA_BRPORT_LEARNING_SYNC])
+ return -EINVAL;
+ enable = !!nla_get_u8(bp_tb[IFLA_BRPORT_LEARNING_SYNC]);
+
+ if (enable == !!(priv->brport_features & BR_LEARNING_SYNC))
+ return 0;
+
+ mutex_lock(&card->sbp_lock);
+ /* do not change anything if BridgePort is enabled */
+ if (qeth_bridgeport_is_in_use(card)) {
+ NL_SET_ERR_MSG(extack, "n/a (BridgePort)");
+ rc = -EBUSY;
+ } else if (enable) {
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_ADDR_INFO);
+ rc = qeth_l2_dev2br_an_set(card, true);
+ if (rc)
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
+ else
+ priv->brport_features |= BR_LEARNING_SYNC;
+ } else {
+ rc = qeth_l2_dev2br_an_set(card, false);
+ if (!rc) {
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
+ priv->brport_features ^= BR_LEARNING_SYNC;
+ qeth_l2_dev2br_fdb_flush(card);
+ }
+ }
+ mutex_unlock(&card->sbp_lock);
+
+ return rc;
+}
+
static const struct net_device_ops qeth_l2_netdev_ops = {
.ndo_open = qeth_open,
.ndo_stop = qeth_stop,
@@ -873,6 +964,7 @@ static const struct net_device_ops qeth_l2_netdev_ops = {
.ndo_fix_features = qeth_fix_features,
.ndo_set_features = qeth_set_features,
.ndo_bridge_getlink = qeth_l2_bridge_getlink,
+ .ndo_bridge_setlink = qeth_l2_bridge_setlink,
};
static const struct net_device_ops qeth_osn_netdev_ops = {
@@ -1016,6 +1108,38 @@ static void qeth_l2_detect_dev2br_support(struct qeth_card *card)
priv->brport_hw_features &= ~BR_LEARNING_SYNC;
}
+static void qeth_l2_enable_brport_features(struct qeth_card *card)
+{
+ struct qeth_priv *priv = netdev_priv(card->dev);
+ int rc;
+
+ if (priv->brport_features & BR_LEARNING_SYNC) {
+ if (priv->brport_hw_features & BR_LEARNING_SYNC) {
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_ADDR_INFO);
+ rc = qeth_l2_dev2br_an_set(card, true);
+ if (rc == -EAGAIN) {
+ /* Recoverable error, retry once */
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
+ qeth_l2_dev2br_fdb_flush(card);
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_ADDR_INFO);
+ rc = qeth_l2_dev2br_an_set(card, true);
+ }
+ if (rc) {
+ netdev_err(card->dev,
+ "failed to enable bridge learning_sync: %d\n",
+ rc);
+ qeth_l2_set_pnso_mode(card, QETH_PNSO_NONE);
+ qeth_l2_dev2br_fdb_flush(card);
+ priv->brport_features ^= BR_LEARNING_SYNC;
+ }
+ } else {
+ dev_warn(&card->gdev->dev,
+ "bridge learning_sync not supported\n");
+ priv->brport_features ^= BR_LEARNING_SYNC;
+ }
+ }
+}
+
static int qeth_l2_set_online(struct qeth_card *card)
{
struct ccwgroup_device *gdev = card->gdev;
@@ -1075,6 +1199,7 @@ static int qeth_l2_set_online(struct qeth_card *card)
netif_device_attach(dev);
qeth_enable_hw_features(dev);
+ qeth_l2_enable_brport_features(card);
if (card->info.open_when_online) {
card->info.open_when_online = 0;
--
2.17.1
^ permalink raw reply related [flat|nested] 13+ messages in thread* Re: [PATCH net-next 0/8] s390/qeth: updates 2020-09-10
2020-09-10 17:23 [PATCH net-next 0/8] s390/qeth: updates 2020-09-10 Julian Wiedmann
` (7 preceding siblings ...)
2020-09-10 17:23 ` [PATCH net-next 8/8] s390/qeth: implement ndo_bridge_setlink " Julian Wiedmann
@ 2020-09-15 20:22 ` David Miller
8 siblings, 0 replies; 13+ messages in thread
From: David Miller @ 2020-09-15 20:22 UTC (permalink / raw)
To: jwi
Cc: kuba, netdev, linux-s390, hca, ubraun, kgraul, wintera, roopa,
nikolay, jiri, ivecera
From: Julian Wiedmann <jwi@linux.ibm.com>
Date: Thu, 10 Sep 2020 19:23:43 +0200
> subject to positive review by the bridge maintainers on patch 5,
> please apply the following patch series to netdev's net-next tree.
>
> Alexandra adds BR_LEARNING_SYNC support to qeth. In addition to the
> main qeth changes (controlling the feature, and raising switchdev
> events), this also needs
> - Patch 1 and 2 for some s390/cio infrastructure improvements
> (acked by Heiko to go in via net-next), and
> - Patch 5 to introduce a new switchdev_notifier_type, so that a driver
> can clear all previously learned entries from the bridge FDB in case
> things go out-of-sync later on.
Series applied, thank you.
^ permalink raw reply [flat|nested] 13+ messages in thread