* [PATCH RESEND net-next 03/12] s390/qeth: display warning for OSA3 RX/TX checksum offloading
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Thomas Richter <tmricht@linux.vnet.ibm.com>
When RX/TX checksum offloading is turned on and the adapter is
an OSA 3 card in layer 3 mode, the checksum offloading is only
performed when both peers use different adapters. If both peers
share an OSA 3 card, communication is a memory copy and
checksum offloading is not performed.
This patch adds a warning to inform the administrator.
OSA 3 in layer 2 mode does not offer the RX/TX checksum
offload feature.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_core_main.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c
index 49b813f..ca8309f 100644
--- a/drivers/s390/net/qeth_core_main.c
+++ b/drivers/s390/net/qeth_core_main.c
@@ -6116,6 +6116,11 @@ static int qeth_send_checksum_on(struct qeth_card *card, int cstype)
if ((required_features & chksum_cb.supported) !=
required_features)
rc = -EIO;
+ else if (!(QETH_IPA_CHECKSUM_LP2LP & chksum_cb.supported) &&
+ cstype == IPA_INBOUND_CHECKSUM)
+ dev_warn(&card->gdev->dev,
+ "Hardware checksumming is performed only if %s and its peer use different OSA Express 3 ports\n",
+ QETH_CARD_IFNAME(card));
}
if (rc) {
qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_STOP, 0);
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 02/12] s390/qeth: test RX/TX checksum offload reply
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Thomas Richter <tmricht@linux.vnet.ibm.com>
Turning on receive and/or transmit checksum offload support
on the OSA card requires 2 commands:
1. start command which replies with available features
2. enable command to turn on selected features.
The current version does not check the reply of the start
command and simply uses the returned value to enable
offload features. When the start command returns zero, this
leads to a situation where no checksum offload
is turned on by the hardware. Even worse no error
indication is returned. The Linux kernel assumes
the OSA card performs RX/TX checksum offload, but the hardware
does not perform any checksum verification at all.
This patch checks the return of the start and enable
command responses from the hardware and turns off
checksum offloading if the commands fails or does not
respond with the correct bit setting.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_core_main.c | 13 +++++++++++++
drivers/s390/net/qeth_core_mpc.h | 10 ++++++++++
2 files changed, 23 insertions(+)
diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c
index 5ab80ea..49b813f 100644
--- a/drivers/s390/net/qeth_core_main.c
+++ b/drivers/s390/net/qeth_core_main.c
@@ -6104,11 +6104,19 @@ static int qeth_ipa_checksum_run_cmd(struct qeth_card *card,
static int qeth_send_checksum_on(struct qeth_card *card, int cstype)
{
+ const __u32 required_features = QETH_IPA_CHECKSUM_IP_HDR |
+ QETH_IPA_CHECKSUM_UDP |
+ QETH_IPA_CHECKSUM_TCP;
struct qeth_checksum_cmd chksum_cb;
int rc;
rc = qeth_ipa_checksum_run_cmd(card, cstype, IPA_CMD_ASS_START, 0,
&chksum_cb);
+ if (!rc) {
+ if ((required_features & chksum_cb.supported) !=
+ required_features)
+ rc = -EIO;
+ }
if (rc) {
qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_STOP, 0);
dev_warn(&card->gdev->dev,
@@ -6118,6 +6126,11 @@ static int qeth_send_checksum_on(struct qeth_card *card, int cstype)
}
rc = qeth_ipa_checksum_run_cmd(card, cstype, IPA_CMD_ASS_ENABLE,
chksum_cb.supported, &chksum_cb);
+ if (!rc) {
+ if ((required_features & chksum_cb.enabled) !=
+ required_features)
+ rc = -EIO;
+ }
if (rc) {
qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_STOP, 0);
dev_warn(&card->gdev->dev,
diff --git a/drivers/s390/net/qeth_core_mpc.h b/drivers/s390/net/qeth_core_mpc.h
index f54ea72..bc69d0a 100644
--- a/drivers/s390/net/qeth_core_mpc.h
+++ b/drivers/s390/net/qeth_core_mpc.h
@@ -352,6 +352,16 @@ struct qeth_arp_query_info {
char *udata;
};
+/* IPA set assist segmentation bit definitions for receive and
+ * transmit checksum offloading.
+ */
+enum qeth_ipa_checksum_bits {
+ QETH_IPA_CHECKSUM_IP_HDR = 0x0002,
+ QETH_IPA_CHECKSUM_UDP = 0x0008,
+ QETH_IPA_CHECKSUM_TCP = 0x0010,
+ QETH_IPA_CHECKSUM_LP2LP = 0x0020
+};
+
/* IPA Assist checksum offload reply layout. */
struct qeth_checksum_cmd {
__u32 supported;
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 00/12] s390: qeth patches
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
Hi Dave,
yesterday I came up with 13 qeth patches. Since you have not been
happy with the 13th patch, I want to make sure that at least the
remaining 12 qeth patches can be applied to net-next. Here is the
resend of them.
Thanks,
Ursula
Julian Wiedmann (8):
s390/qeth: Allow reading hsuid in state DOWN
s390/qeth: Remove QETH_IP_HEADER_SIZE
s390/qeth: drop qeth_l2_del_all_macs() parameter
s390/qeth: don't convert return code twice
s390/qeth: consolidate errno translation
s390/qeth: extract qeth_l2_remove_mac()
s390/qeth: shuffle MAC management functions around
s390/qeth: issue STARTLAN as first IPA command
Thomas Richter (3):
s390/qeth: rework RX/TX checksum offload
s390/qeth: test RX/TX checksum offload reply
s390/qeth: display warning for OSA3 RX/TX checksum offloading
Ursula Braun (1):
s390/qeth: fix retrieval of vipa and proxy-arp addresses
drivers/s390/net/qeth_core.h | 5 -
drivers/s390/net/qeth_core_main.c | 135 ++++++++++++++++++++-------
drivers/s390/net/qeth_core_mpc.h | 17 ++++
drivers/s390/net/qeth_l2_main.c | 189 ++++++++++++++++----------------------
drivers/s390/net/qeth_l3_main.c | 15 ---
drivers/s390/net/qeth_l3_sys.c | 33 ++++---
6 files changed, 212 insertions(+), 182 deletions(-)
--
2.8.4
^ permalink raw reply
* [PATCH RESEND net-next 10/12] s390/qeth: shuffle MAC management functions around
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Move all MAC utility functions in one place, and drop the
forward declarations.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 129 ++++++++++++++++++++--------------------
1 file changed, 63 insertions(+), 66 deletions(-)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index d456740..c298759c 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -27,9 +27,6 @@
static int qeth_l2_set_offline(struct ccwgroup_device *);
static int qeth_l2_stop(struct net_device *);
-static int qeth_l2_send_delmac(struct qeth_card *, __u8 *);
-static int qeth_l2_send_setdelmac(struct qeth_card *, __u8 *,
- enum qeth_ipa_cmds);
static void qeth_l2_set_rx_mode(struct net_device *);
static int qeth_l2_recover(void *);
static void qeth_bridgeport_query_support(struct qeth_card *card);
@@ -165,6 +162,64 @@ static int qeth_setdel_makerc(struct qeth_card *card, int retcode)
return rc;
}
+static int qeth_l2_send_setdelmac(struct qeth_card *card, __u8 *mac,
+ enum qeth_ipa_cmds ipacmd)
+{
+ struct qeth_ipa_cmd *cmd;
+ struct qeth_cmd_buffer *iob;
+
+ QETH_CARD_TEXT(card, 2, "L2sdmac");
+ iob = qeth_get_ipacmd_buffer(card, ipacmd, QETH_PROT_IPV4);
+ if (!iob)
+ return -ENOMEM;
+ cmd = (struct qeth_ipa_cmd *)(iob->data+IPA_PDU_HEADER_SIZE);
+ cmd->data.setdelmac.mac_length = OSA_ADDR_LEN;
+ memcpy(&cmd->data.setdelmac.mac, mac, OSA_ADDR_LEN);
+ return qeth_setdel_makerc(card, qeth_send_ipa_cmd(card, iob,
+ NULL, NULL));
+}
+
+static int qeth_l2_send_setmac(struct qeth_card *card, __u8 *mac)
+{
+ int rc;
+
+ QETH_CARD_TEXT(card, 2, "L2Setmac");
+ rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_SETVMAC);
+ if (rc == 0) {
+ card->info.mac_bits |= QETH_LAYER2_MAC_REGISTERED;
+ memcpy(card->dev->dev_addr, mac, OSA_ADDR_LEN);
+ dev_info(&card->gdev->dev,
+ "MAC address %pM successfully registered on device %s\n",
+ card->dev->dev_addr, card->dev->name);
+ } else {
+ card->info.mac_bits &= ~QETH_LAYER2_MAC_REGISTERED;
+ switch (rc) {
+ case -EEXIST:
+ dev_warn(&card->gdev->dev,
+ "MAC address %pM already exists\n", mac);
+ break;
+ case -EPERM:
+ dev_warn(&card->gdev->dev,
+ "MAC address %pM is not authorized\n", mac);
+ break;
+ }
+ }
+ return rc;
+}
+
+static int qeth_l2_send_delmac(struct qeth_card *card, __u8 *mac)
+{
+ int rc;
+
+ QETH_CARD_TEXT(card, 2, "L2Delmac");
+ if (!(card->info.mac_bits & QETH_LAYER2_MAC_REGISTERED))
+ return 0;
+ rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_DELVMAC);
+ if (rc == 0)
+ card->info.mac_bits &= ~QETH_LAYER2_MAC_REGISTERED;
+ return rc;
+}
+
static int qeth_l2_send_setgroupmac(struct qeth_card *card, __u8 *mac)
{
int rc;
@@ -193,11 +248,6 @@ static int qeth_l2_send_delgroupmac(struct qeth_card *card, __u8 *mac)
return rc;
}
-static inline u32 qeth_l2_mac_hash(const u8 *addr)
-{
- return get_unaligned((u32 *)(&addr[2]));
-}
-
static int qeth_l2_write_mac(struct qeth_card *card, struct qeth_mac *mac)
{
if (mac->is_uc) {
@@ -232,6 +282,11 @@ static void qeth_l2_del_all_macs(struct qeth_card *card)
spin_unlock_bh(&card->mclock);
}
+static inline u32 qeth_l2_mac_hash(const u8 *addr)
+{
+ return get_unaligned((u32 *)(&addr[2]));
+}
+
static inline int qeth_l2_get_cast_type(struct qeth_card *card,
struct sk_buff *skb)
{
@@ -572,64 +627,6 @@ static int qeth_l2_poll(struct napi_struct *napi, int budget)
return work_done;
}
-static int qeth_l2_send_setdelmac(struct qeth_card *card, __u8 *mac,
- enum qeth_ipa_cmds ipacmd)
-{
- struct qeth_ipa_cmd *cmd;
- struct qeth_cmd_buffer *iob;
-
- QETH_CARD_TEXT(card, 2, "L2sdmac");
- iob = qeth_get_ipacmd_buffer(card, ipacmd, QETH_PROT_IPV4);
- if (!iob)
- return -ENOMEM;
- cmd = (struct qeth_ipa_cmd *)(iob->data+IPA_PDU_HEADER_SIZE);
- cmd->data.setdelmac.mac_length = OSA_ADDR_LEN;
- memcpy(&cmd->data.setdelmac.mac, mac, OSA_ADDR_LEN);
- return qeth_setdel_makerc(card, qeth_send_ipa_cmd(card, iob,
- NULL, NULL));
-}
-
-static int qeth_l2_send_setmac(struct qeth_card *card, __u8 *mac)
-{
- int rc;
-
- QETH_CARD_TEXT(card, 2, "L2Setmac");
- rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_SETVMAC);
- if (rc == 0) {
- card->info.mac_bits |= QETH_LAYER2_MAC_REGISTERED;
- memcpy(card->dev->dev_addr, mac, OSA_ADDR_LEN);
- dev_info(&card->gdev->dev,
- "MAC address %pM successfully registered on device %s\n",
- card->dev->dev_addr, card->dev->name);
- } else {
- card->info.mac_bits &= ~QETH_LAYER2_MAC_REGISTERED;
- switch (rc) {
- case -EEXIST:
- dev_warn(&card->gdev->dev,
- "MAC address %pM already exists\n", mac);
- break;
- case -EPERM:
- dev_warn(&card->gdev->dev,
- "MAC address %pM is not authorized\n", mac);
- break;
- }
- }
- return rc;
-}
-
-static int qeth_l2_send_delmac(struct qeth_card *card, __u8 *mac)
-{
- int rc;
-
- QETH_CARD_TEXT(card, 2, "L2Delmac");
- if (!(card->info.mac_bits & QETH_LAYER2_MAC_REGISTERED))
- return 0;
- rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_DELVMAC);
- if (rc == 0)
- card->info.mac_bits &= ~QETH_LAYER2_MAC_REGISTERED;
- return rc;
-}
-
static int qeth_l2_request_initial_mac(struct qeth_card *card)
{
int rc = 0;
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 07/12] s390/qeth: don't convert return code twice
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
qeth_l2_send_groupmac() already translates the return code, so
calling qeth_setdel_makerc() a second time only produces garbage.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 3025f56..38fae10 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -210,8 +210,7 @@ static int qeth_l2_write_mac(struct qeth_card *card, struct qeth_mac *mac)
qeth_l2_send_setdelmac(card, mac->mac_addr,
IPA_CMD_SETVMAC));
} else {
- rc = qeth_setdel_makerc(card,
- qeth_l2_send_setgroupmac(card, mac->mac_addr));
+ rc = qeth_l2_send_setgroupmac(card, mac->mac_addr);
}
return rc;
}
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 01/12] s390/qeth: rework RX/TX checksum offload
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Thomas Richter <tmricht@linux.vnet.ibm.com>
Rework the RX/TX checksum offloading command sequence to use
the provided function call back mechanims to return card
data to the device driver.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_core.h | 2 -
drivers/s390/net/qeth_core_main.c | 96 ++++++++++++++++++++++++++-------------
drivers/s390/net/qeth_core_mpc.h | 7 +++
3 files changed, 72 insertions(+), 33 deletions(-)
diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h
index 6d4b68c4..41e4665 100644
--- a/drivers/s390/net/qeth_core.h
+++ b/drivers/s390/net/qeth_core.h
@@ -674,8 +674,6 @@ struct qeth_card_info {
int broadcast_capable;
int unique_id;
struct qeth_card_blkt blkt;
- __u32 csum_mask;
- __u32 tx_csum_mask;
enum qeth_ipa_promisc_modes promisc_mode;
__u32 diagass_support;
__u32 hwtrap;
diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c
index e335583..5ab80ea 100644
--- a/drivers/s390/net/qeth_core_main.c
+++ b/drivers/s390/net/qeth_core_main.c
@@ -5289,18 +5289,6 @@ int qeth_setassparms_cb(struct qeth_card *card,
if (cmd->hdr.prot_version == QETH_PROT_IPV6)
card->options.ipa6.enabled_funcs = cmd->hdr.ipa_enabled;
}
- if (cmd->data.setassparms.hdr.assist_no == IPA_INBOUND_CHECKSUM &&
- cmd->data.setassparms.hdr.command_code == IPA_CMD_ASS_START) {
- card->info.csum_mask = cmd->data.setassparms.data.flags_32bit;
- QETH_CARD_TEXT_(card, 3, "csum:%d", card->info.csum_mask);
- }
- if (cmd->data.setassparms.hdr.assist_no == IPA_OUTBOUND_CHECKSUM &&
- cmd->data.setassparms.hdr.command_code == IPA_CMD_ASS_START) {
- card->info.tx_csum_mask =
- cmd->data.setassparms.data.flags_32bit;
- QETH_CARD_TEXT_(card, 3, "tcsu:%d", card->info.tx_csum_mask);
- }
-
return 0;
}
EXPORT_SYMBOL_GPL(qeth_setassparms_cb);
@@ -6060,23 +6048,78 @@ int qeth_core_ethtool_get_settings(struct net_device *netdev,
}
EXPORT_SYMBOL_GPL(qeth_core_ethtool_get_settings);
+/* Callback to handle checksum offload command reply from OSA card.
+ * Verify that required features have been enabled on the card.
+ * Return error in hdr->return_code as this value is checked by caller.
+ *
+ * Always returns zero to indicate no further messages from the OSA card.
+ */
+static int qeth_ipa_checksum_run_cmd_cb(struct qeth_card *card,
+ struct qeth_reply *reply,
+ unsigned long data)
+{
+ struct qeth_ipa_cmd *cmd = (struct qeth_ipa_cmd *) data;
+ struct qeth_checksum_cmd *chksum_cb =
+ (struct qeth_checksum_cmd *)reply->param;
+
+ QETH_CARD_TEXT(card, 4, "chkdoccb");
+ if (cmd->hdr.return_code)
+ return 0;
+
+ memset(chksum_cb, 0, sizeof(*chksum_cb));
+ if (cmd->data.setassparms.hdr.command_code == IPA_CMD_ASS_START) {
+ chksum_cb->supported =
+ cmd->data.setassparms.data.chksum.supported;
+ QETH_CARD_TEXT_(card, 3, "strt:%x", chksum_cb->supported);
+ }
+ if (cmd->data.setassparms.hdr.command_code == IPA_CMD_ASS_ENABLE) {
+ chksum_cb->supported =
+ cmd->data.setassparms.data.chksum.supported;
+ chksum_cb->enabled =
+ cmd->data.setassparms.data.chksum.enabled;
+ QETH_CARD_TEXT_(card, 3, "supp:%x", chksum_cb->supported);
+ QETH_CARD_TEXT_(card, 3, "enab:%x", chksum_cb->enabled);
+ }
+ return 0;
+}
+
+/* Send command to OSA card and check results. */
+static int qeth_ipa_checksum_run_cmd(struct qeth_card *card,
+ enum qeth_ipa_funcs ipa_func,
+ __u16 cmd_code, long data,
+ struct qeth_checksum_cmd *chksum_cb)
+{
+ struct qeth_cmd_buffer *iob;
+ int rc = -ENOMEM;
+
+ QETH_CARD_TEXT(card, 4, "chkdocmd");
+ iob = qeth_get_setassparms_cmd(card, ipa_func, cmd_code,
+ sizeof(__u32), QETH_PROT_IPV4);
+ if (iob)
+ rc = qeth_send_setassparms(card, iob, sizeof(__u32), data,
+ qeth_ipa_checksum_run_cmd_cb,
+ chksum_cb);
+ return rc;
+}
+
static int qeth_send_checksum_on(struct qeth_card *card, int cstype)
{
- long rxtx_arg;
+ struct qeth_checksum_cmd chksum_cb;
int rc;
- rc = qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_START, 0);
+ rc = qeth_ipa_checksum_run_cmd(card, cstype, IPA_CMD_ASS_START, 0,
+ &chksum_cb);
if (rc) {
+ qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_STOP, 0);
dev_warn(&card->gdev->dev,
"Starting HW checksumming for %s failed, using SW checksumming\n",
QETH_CARD_IFNAME(card));
return rc;
}
- rxtx_arg = (cstype == IPA_OUTBOUND_CHECKSUM) ? card->info.tx_csum_mask
- : card->info.csum_mask;
- rc = qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_ENABLE,
- rxtx_arg);
+ rc = qeth_ipa_checksum_run_cmd(card, cstype, IPA_CMD_ASS_ENABLE,
+ chksum_cb.supported, &chksum_cb);
if (rc) {
+ qeth_send_simple_setassparms(card, cstype, IPA_CMD_ASS_STOP, 0);
dev_warn(&card->gdev->dev,
"Enabling HW checksumming for %s failed, using SW checksumming\n",
QETH_CARD_IFNAME(card));
@@ -6090,19 +6133,10 @@ static int qeth_send_checksum_on(struct qeth_card *card, int cstype)
static int qeth_set_ipa_csum(struct qeth_card *card, int on, int cstype)
{
- int rc;
-
- if (on) {
- rc = qeth_send_checksum_on(card, cstype);
- if (rc)
- return -EIO;
- } else {
- rc = qeth_send_simple_setassparms(card, cstype,
- IPA_CMD_ASS_STOP, 0);
- if (rc)
- return -EIO;
- }
- return 0;
+ int rc = (on) ? qeth_send_checksum_on(card, cstype)
+ : qeth_send_simple_setassparms(card, cstype,
+ IPA_CMD_ASS_STOP, 0);
+ return rc ? -EIO : 0;
}
static int qeth_set_ipa_tso(struct qeth_card *card, int on)
diff --git a/drivers/s390/net/qeth_core_mpc.h b/drivers/s390/net/qeth_core_mpc.h
index 6cccc9a..f54ea72 100644
--- a/drivers/s390/net/qeth_core_mpc.h
+++ b/drivers/s390/net/qeth_core_mpc.h
@@ -352,11 +352,18 @@ struct qeth_arp_query_info {
char *udata;
};
+/* IPA Assist checksum offload reply layout. */
+struct qeth_checksum_cmd {
+ __u32 supported;
+ __u32 enabled;
+} __packed;
+
/* SETASSPARMS IPA Command: */
struct qeth_ipacmd_setassparms {
struct qeth_ipacmd_setassparms_hdr hdr;
union {
__u32 flags_32bit;
+ struct qeth_checksum_cmd chksum;
struct qeth_arp_cache_entry add_arp_entry;
struct qeth_arp_query_data query_arp;
__u8 ip[16];
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 08/12] s390/qeth: consolidate errno translation
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Consolidate errno handling for MAC management: Instead of doing this in every
caller, do it in one place.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Suggested-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 20 ++++++++------------
1 file changed, 8 insertions(+), 12 deletions(-)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 38fae10..074fc62 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -170,8 +170,7 @@ static int qeth_l2_send_setgroupmac(struct qeth_card *card, __u8 *mac)
int rc;
QETH_CARD_TEXT(card, 2, "L2Sgmac");
- rc = qeth_setdel_makerc(card, qeth_l2_send_setdelmac(card, mac,
- IPA_CMD_SETGMAC));
+ rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_SETGMAC);
if (rc == -EEXIST)
QETH_DBF_MESSAGE(2, "Group MAC %pM already existing on %s\n",
mac, QETH_CARD_IFNAME(card));
@@ -186,8 +185,7 @@ static int qeth_l2_send_delgroupmac(struct qeth_card *card, __u8 *mac)
int rc;
QETH_CARD_TEXT(card, 2, "L2Dgmac");
- rc = qeth_setdel_makerc(card, qeth_l2_send_setdelmac(card, mac,
- IPA_CMD_DELGMAC));
+ rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_DELGMAC);
if (rc)
QETH_DBF_MESSAGE(2,
"Could not delete group MAC %pM on %s: %d\n",
@@ -206,9 +204,8 @@ static int qeth_l2_write_mac(struct qeth_card *card, struct qeth_mac *mac)
int rc;
if (mac->is_uc) {
- rc = qeth_setdel_makerc(card,
- qeth_l2_send_setdelmac(card, mac->mac_addr,
- IPA_CMD_SETVMAC));
+ rc = qeth_l2_send_setdelmac(card, mac->mac_addr,
+ IPA_CMD_SETVMAC);
} else {
rc = qeth_l2_send_setgroupmac(card, mac->mac_addr);
}
@@ -582,7 +579,8 @@ static int qeth_l2_send_setdelmac(struct qeth_card *card, __u8 *mac,
cmd = (struct qeth_ipa_cmd *)(iob->data+IPA_PDU_HEADER_SIZE);
cmd->data.setdelmac.mac_length = OSA_ADDR_LEN;
memcpy(&cmd->data.setdelmac.mac, mac, OSA_ADDR_LEN);
- return qeth_send_ipa_cmd(card, iob, NULL, NULL);
+ return qeth_setdel_makerc(card, qeth_send_ipa_cmd(card, iob,
+ NULL, NULL));
}
static int qeth_l2_send_setmac(struct qeth_card *card, __u8 *mac)
@@ -590,8 +588,7 @@ static int qeth_l2_send_setmac(struct qeth_card *card, __u8 *mac)
int rc;
QETH_CARD_TEXT(card, 2, "L2Setmac");
- rc = qeth_setdel_makerc(card, qeth_l2_send_setdelmac(card, mac,
- IPA_CMD_SETVMAC));
+ rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_SETVMAC);
if (rc == 0) {
card->info.mac_bits |= QETH_LAYER2_MAC_REGISTERED;
memcpy(card->dev->dev_addr, mac, OSA_ADDR_LEN);
@@ -621,8 +618,7 @@ static int qeth_l2_send_delmac(struct qeth_card *card, __u8 *mac)
QETH_CARD_TEXT(card, 2, "L2Delmac");
if (!(card->info.mac_bits & QETH_LAYER2_MAC_REGISTERED))
return 0;
- rc = qeth_setdel_makerc(card, qeth_l2_send_setdelmac(card, mac,
- IPA_CMD_DELVMAC));
+ rc = qeth_l2_send_setdelmac(card, mac, IPA_CMD_DELVMAC);
if (rc == 0)
card->info.mac_bits &= ~QETH_LAYER2_MAC_REGISTERED;
return rc;
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 09/12] s390/qeth: extract qeth_l2_remove_mac()
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
This matches qeth_l2_write_mac().
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 27 +++++++++++++--------------
1 file changed, 13 insertions(+), 14 deletions(-)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 074fc62..d456740 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -200,16 +200,22 @@ static inline u32 qeth_l2_mac_hash(const u8 *addr)
static int qeth_l2_write_mac(struct qeth_card *card, struct qeth_mac *mac)
{
-
- int rc;
-
if (mac->is_uc) {
- rc = qeth_l2_send_setdelmac(card, mac->mac_addr,
+ return qeth_l2_send_setdelmac(card, mac->mac_addr,
IPA_CMD_SETVMAC);
} else {
- rc = qeth_l2_send_setgroupmac(card, mac->mac_addr);
+ return qeth_l2_send_setgroupmac(card, mac->mac_addr);
+ }
+}
+
+static int qeth_l2_remove_mac(struct qeth_card *card, struct qeth_mac *mac)
+{
+ if (mac->is_uc) {
+ return qeth_l2_send_setdelmac(card, mac->mac_addr,
+ IPA_CMD_DELVMAC);
+ } else {
+ return qeth_l2_send_delgroupmac(card, mac->mac_addr);
}
- return rc;
}
static void qeth_l2_del_all_macs(struct qeth_card *card)
@@ -782,14 +788,7 @@ static void qeth_l2_set_rx_mode(struct net_device *dev)
hash_for_each_safe(card->mac_htable, i, tmp, mac, hnode) {
if (mac->disp_flag == QETH_DISP_ADDR_DELETE) {
- if (!mac->is_uc)
- rc = qeth_l2_send_delgroupmac(card,
- mac->mac_addr);
- else {
- rc = qeth_l2_send_setdelmac(card, mac->mac_addr,
- IPA_CMD_DELVMAC);
- }
-
+ qeth_l2_remove_mac(card, mac);
hash_del(&mac->hnode);
kfree(mac);
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 04/12] s390/qeth: Allow reading hsuid in state DOWN
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Accessing the current hsuid via card->options.hsuid is perfectly
fine, even when the card is DOWN.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_l3_sys.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/s390/net/qeth_l3_sys.c b/drivers/s390/net/qeth_l3_sys.c
index 0e00a5c..3cd4d9f 100644
--- a/drivers/s390/net/qeth_l3_sys.c
+++ b/drivers/s390/net/qeth_l3_sys.c
@@ -250,9 +250,6 @@ static ssize_t qeth_l3_dev_hsuid_show(struct device *dev,
if (card->info.type != QETH_CARD_TYPE_IQD)
return -EPERM;
- if (card->state == CARD_STATE_DOWN)
- return -EPERM;
-
memcpy(tmp_hsuid, card->options.hsuid, sizeof(tmp_hsuid));
EBCASC(tmp_hsuid, 8);
return sprintf(buf, "%s\n", tmp_hsuid);
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 11/12] s390/qeth: issue STARTLAN as first IPA command
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
STARTLAN needs to be the first IPA command after MPC initialization
completes.
So move the qeth_send_startlan() call from the layer disciplines
into the core path, right after the MPC handshake.
While at it, replace the magic LAN OFFLINE return code
with the existing enum.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_core.h | 1 -
drivers/s390/net/qeth_core_main.c | 21 +++++++++++++++++----
drivers/s390/net/qeth_l2_main.c | 15 ---------------
drivers/s390/net/qeth_l3_main.c | 15 ---------------
4 files changed, 17 insertions(+), 35 deletions(-)
diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h
index 774ae51..e7addea 100644
--- a/drivers/s390/net/qeth_core.h
+++ b/drivers/s390/net/qeth_core.h
@@ -913,7 +913,6 @@ void qeth_clear_thread_running_bit(struct qeth_card *, unsigned long);
int qeth_core_hardsetup_card(struct qeth_card *);
void qeth_print_status_message(struct qeth_card *);
int qeth_init_qdio_queues(struct qeth_card *);
-int qeth_send_startlan(struct qeth_card *);
int qeth_send_ipa_cmd(struct qeth_card *, struct qeth_cmd_buffer *,
int (*reply_cb)
(struct qeth_card *, struct qeth_reply *, unsigned long),
diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c
index ca8309f..315d8a2 100644
--- a/drivers/s390/net/qeth_core_main.c
+++ b/drivers/s390/net/qeth_core_main.c
@@ -2944,7 +2944,7 @@ int qeth_send_ipa_cmd(struct qeth_card *card, struct qeth_cmd_buffer *iob,
}
EXPORT_SYMBOL_GPL(qeth_send_ipa_cmd);
-int qeth_send_startlan(struct qeth_card *card)
+static int qeth_send_startlan(struct qeth_card *card)
{
int rc;
struct qeth_cmd_buffer *iob;
@@ -2957,7 +2957,6 @@ int qeth_send_startlan(struct qeth_card *card)
rc = qeth_send_ipa_cmd(card, iob, NULL, NULL);
return rc;
}
-EXPORT_SYMBOL_GPL(qeth_send_startlan);
static int qeth_default_setadapterparms_cb(struct qeth_card *card,
struct qeth_reply *reply, unsigned long data)
@@ -5087,6 +5086,20 @@ int qeth_core_hardsetup_card(struct qeth_card *card)
goto out;
}
+ rc = qeth_send_startlan(card);
+ if (rc) {
+ QETH_DBF_TEXT_(SETUP, 2, "6err%d", rc);
+ if (rc == IPA_RC_LAN_OFFLINE) {
+ dev_warn(&card->gdev->dev,
+ "The LAN is offline\n");
+ card->lan_online = 0;
+ } else {
+ rc = -ENODEV;
+ goto out;
+ }
+ } else
+ card->lan_online = 1;
+
card->options.ipa4.supported_funcs = 0;
card->options.ipa6.supported_funcs = 0;
card->options.adp.supported_funcs = 0;
@@ -5098,14 +5111,14 @@ int qeth_core_hardsetup_card(struct qeth_card *card)
if (qeth_is_supported(card, IPA_SETADAPTERPARMS)) {
rc = qeth_query_setadapterparms(card);
if (rc < 0) {
- QETH_DBF_TEXT_(SETUP, 2, "6err%d", rc);
+ QETH_DBF_TEXT_(SETUP, 2, "7err%d", rc);
goto out;
}
}
if (qeth_adp_supported(card, IPA_SETADP_SET_DIAG_ASSIST)) {
rc = qeth_query_setdiagass(card);
if (rc < 0) {
- QETH_DBF_TEXT_(SETUP, 2, "7err%d", rc);
+ QETH_DBF_TEXT_(SETUP, 2, "8err%d", rc);
goto out;
}
}
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index c298759c..bea4833 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -1177,21 +1177,6 @@ static int __qeth_l2_set_online(struct ccwgroup_device *gdev, int recovery_mode)
/* softsetup */
QETH_DBF_TEXT(SETUP, 2, "softsetp");
- rc = qeth_send_startlan(card);
- if (rc) {
- QETH_DBF_TEXT_(SETUP, 2, "1err%d", rc);
- if (rc == 0xe080) {
- dev_warn(&card->gdev->dev,
- "The LAN is offline\n");
- card->lan_online = 0;
- goto contin;
- }
- rc = -ENODEV;
- goto out_remove;
- } else
- card->lan_online = 1;
-
-contin:
if ((card->info.type == QETH_CARD_TYPE_OSD) ||
(card->info.type == QETH_CARD_TYPE_OSX)) {
rc = qeth_l2_start_ipassists(card);
diff --git a/drivers/s390/net/qeth_l3_main.c b/drivers/s390/net/qeth_l3_main.c
index ac37d05..06d0add 100644
--- a/drivers/s390/net/qeth_l3_main.c
+++ b/drivers/s390/net/qeth_l3_main.c
@@ -3227,21 +3227,6 @@ static int __qeth_l3_set_online(struct ccwgroup_device *gdev, int recovery_mode)
/* softsetup */
QETH_DBF_TEXT(SETUP, 2, "softsetp");
- rc = qeth_send_startlan(card);
- if (rc) {
- QETH_DBF_TEXT_(SETUP, 2, "1err%d", rc);
- if (rc == 0xe080) {
- dev_warn(&card->gdev->dev,
- "The LAN is offline\n");
- card->lan_online = 0;
- goto contin;
- }
- rc = -ENODEV;
- goto out_remove;
- } else
- card->lan_online = 1;
-
-contin:
rc = qeth_l3_setadapter_parms(card);
if (rc)
QETH_DBF_TEXT_(SETUP, 2, "2err%04x", rc);
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 06/12] s390/qeth: drop qeth_l2_del_all_macs() parameter
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
The only caller passes del = 0, so remove both the parameter and
the code that handles != 0.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_l2_main.c | 11 ++---------
1 file changed, 2 insertions(+), 9 deletions(-)
diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c
index 9c921c28..3025f56 100644
--- a/drivers/s390/net/qeth_l2_main.c
+++ b/drivers/s390/net/qeth_l2_main.c
@@ -216,7 +216,7 @@ static int qeth_l2_write_mac(struct qeth_card *card, struct qeth_mac *mac)
return rc;
}
-static void qeth_l2_del_all_macs(struct qeth_card *card, int del)
+static void qeth_l2_del_all_macs(struct qeth_card *card)
{
struct qeth_mac *mac;
struct hlist_node *tmp;
@@ -224,13 +224,6 @@ static void qeth_l2_del_all_macs(struct qeth_card *card, int del)
spin_lock_bh(&card->mclock);
hash_for_each_safe(card->mac_htable, i, tmp, mac, hnode) {
- if (del) {
- if (mac->is_uc)
- qeth_l2_send_setdelmac(card, mac->mac_addr,
- IPA_CMD_DELVMAC);
- else
- qeth_l2_send_delgroupmac(card, mac->mac_addr);
- }
hash_del(&mac->hnode);
kfree(mac);
}
@@ -425,7 +418,7 @@ static void qeth_l2_stop_card(struct qeth_card *card, int recovery_mode)
card->state = CARD_STATE_SOFTSETUP;
}
if (card->state == CARD_STATE_SOFTSETUP) {
- qeth_l2_del_all_macs(card, 0);
+ qeth_l2_del_all_macs(card);
qeth_clear_ipacmd_list(card);
card->state = CARD_STATE_HARDSETUP;
}
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 12/12] s390/qeth: fix retrieval of vipa and proxy-arp addresses
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
qeth devices in layer3 mode need a separate handling of vipa and proxy-arp
addresses. vipa and proxy-arp addresses processed by qeth can be read from
userspace. Introduced with commit 5f78e29ceebf ("qeth: optimize IP handling
in rx_mode callback") the retrieval of vipa and proxy-arp addresses is
broken, if more than one vipa or proxy-arp address are set.
The qeth code used local variable "int i" for 2 different purposes. This
patch now spends 2 separate local variables of type "int".
While touching these functions hash_for_each_safe() is converted to
hash_for_each(), since there is no removal of hash entries.
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reference-ID: RQM 3524
---
drivers/s390/net/qeth_l3_sys.c | 30 ++++++++++++++++--------------
1 file changed, 16 insertions(+), 14 deletions(-)
diff --git a/drivers/s390/net/qeth_l3_sys.c b/drivers/s390/net/qeth_l3_sys.c
index 3cd4d9f..05e9471 100644
--- a/drivers/s390/net/qeth_l3_sys.c
+++ b/drivers/s390/net/qeth_l3_sys.c
@@ -689,15 +689,15 @@ static ssize_t qeth_l3_dev_vipa_add_show(char *buf, struct qeth_card *card,
enum qeth_prot_versions proto)
{
struct qeth_ipaddr *ipaddr;
- struct hlist_node *tmp;
char addr_str[40];
+ int str_len = 0;
int entry_len; /* length of 1 entry string, differs between v4 and v6 */
- int i = 0;
+ int i;
entry_len = (proto == QETH_PROT_IPV4)? 12 : 40;
entry_len += 2; /* \n + terminator */
spin_lock_bh(&card->ip_lock);
- hash_for_each_safe(card->ip_htable, i, tmp, ipaddr, hnode) {
+ hash_for_each(card->ip_htable, i, ipaddr, hnode) {
if (ipaddr->proto != proto)
continue;
if (ipaddr->type != QETH_IP_TYPE_VIPA)
@@ -705,16 +705,17 @@ static ssize_t qeth_l3_dev_vipa_add_show(char *buf, struct qeth_card *card,
/* String must not be longer than PAGE_SIZE. So we check if
* string length gets near PAGE_SIZE. Then we can savely display
* the next IPv6 address (worst case, compared to IPv4) */
- if ((PAGE_SIZE - i) <= entry_len)
+ if ((PAGE_SIZE - str_len) <= entry_len)
break;
qeth_l3_ipaddr_to_string(proto, (const u8 *)&ipaddr->u,
addr_str);
- i += snprintf(buf + i, PAGE_SIZE - i, "%s\n", addr_str);
+ str_len += snprintf(buf + str_len, PAGE_SIZE - str_len, "%s\n",
+ addr_str);
}
spin_unlock_bh(&card->ip_lock);
- i += snprintf(buf + i, PAGE_SIZE - i, "\n");
+ str_len += snprintf(buf + str_len, PAGE_SIZE - str_len, "\n");
- return i;
+ return str_len;
}
static ssize_t qeth_l3_dev_vipa_add4_show(struct device *dev,
@@ -851,15 +852,15 @@ static ssize_t qeth_l3_dev_rxip_add_show(char *buf, struct qeth_card *card,
enum qeth_prot_versions proto)
{
struct qeth_ipaddr *ipaddr;
- struct hlist_node *tmp;
char addr_str[40];
+ int str_len = 0;
int entry_len; /* length of 1 entry string, differs between v4 and v6 */
- int i = 0;
+ int i;
entry_len = (proto == QETH_PROT_IPV4)? 12 : 40;
entry_len += 2; /* \n + terminator */
spin_lock_bh(&card->ip_lock);
- hash_for_each_safe(card->ip_htable, i, tmp, ipaddr, hnode) {
+ hash_for_each(card->ip_htable, i, ipaddr, hnode) {
if (ipaddr->proto != proto)
continue;
if (ipaddr->type != QETH_IP_TYPE_RXIP)
@@ -867,16 +868,17 @@ static ssize_t qeth_l3_dev_rxip_add_show(char *buf, struct qeth_card *card,
/* String must not be longer than PAGE_SIZE. So we check if
* string length gets near PAGE_SIZE. Then we can savely display
* the next IPv6 address (worst case, compared to IPv4) */
- if ((PAGE_SIZE - i) <= entry_len)
+ if ((PAGE_SIZE - str_len) <= entry_len)
break;
qeth_l3_ipaddr_to_string(proto, (const u8 *)&ipaddr->u,
addr_str);
- i += snprintf(buf + i, PAGE_SIZE - i, "%s\n", addr_str);
+ str_len += snprintf(buf + str_len, PAGE_SIZE - str_len, "%s\n",
+ addr_str);
}
spin_unlock_bh(&card->ip_lock);
- i += snprintf(buf + i, PAGE_SIZE - i, "\n");
+ str_len += snprintf(buf + str_len, PAGE_SIZE - str_len, "\n");
- return i;
+ return str_len;
}
static ssize_t qeth_l3_dev_rxip_add4_show(struct device *dev,
--
2.8.4
^ permalink raw reply related
* [PATCH RESEND net-next 05/12] s390/qeth: Remove QETH_IP_HEADER_SIZE
From: Ursula Braun @ 2017-01-12 14:48 UTC (permalink / raw)
To: davem; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens, ubraun
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Remove unused define QETH_IP_HEADER_SIZE.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
---
drivers/s390/net/qeth_core.h | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h
index 41e4665..774ae51 100644
--- a/drivers/s390/net/qeth_core.h
+++ b/drivers/s390/net/qeth_core.h
@@ -281,8 +281,6 @@ static inline int qeth_is_ipa_enabled(struct qeth_ipa_info *ipa,
#define QETH_HIGH_WATERMARK_PACK 5
#define QETH_WATERMARK_PACK_FUZZ 1
-#define QETH_IP_HEADER_SIZE 40
-
/* large receive scatter gather copy break */
#define QETH_RX_SG_CB (PAGE_SIZE >> 1)
#define QETH_RX_PULL_LEN 256
--
2.8.4
^ permalink raw reply related
* [PATCH net-next] ipmr: improve hash scalability
From: Nikolay Aleksandrov @ 2017-01-12 14:53 UTC (permalink / raw)
To: netdev; +Cc: roopa, sharpd, davem, Nikolay Aleksandrov
Recently we started using ipmr with thousands of entries and easily hit
soft lockups on smaller devices. The reason is that the hash function
uses the high order bits from the src and dst, but those don't change in
many common cases, also the hash table is only 64 elements so with
thousands it doesn't scale at all.
This patch migrates the hash table to rhashtable, and in particular the
rhl interface which allows for duplicate elements to be chained because
of the MFC_PROXY support (*,G; *,*,oif cases) which allows for multiple
duplicate entries to be added with different interfaces (IMO wrong, but
it's been in for a long time).
And here are some results from tests I've run in a VM:
mr_table size (default, allocated for all namespaces):
Before After
49304 bytes 2400 bytes
Add 65000 routes (the diff is much larger on smaller devices):
Before After
1m42s 58s
Forwarding 256 byte packets with 65000 routes (test done in a VM):
Before After
3 Mbps / ~1465 pps 122 Mbps / ~59000 pps
As a bonus we no longer see the soft lockups on smaller devices which
showed up even with 2000 entries before.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
---
include/linux/mroute.h | 57 ++++++++---
net/ipv4/ipmr.c | 255 +++++++++++++++++++++++++++----------------------
2 files changed, 182 insertions(+), 130 deletions(-)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index f019b62f27b5..d7f63339ef0b 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -3,6 +3,7 @@
#include <linux/in.h>
#include <linux/pim.h>
+#include <linux/rhashtable.h>
#include <net/sock.h>
#include <uapi/linux/mroute.h>
@@ -60,7 +61,6 @@ struct vif_device {
#define VIFF_STATIC 0x8000
#define VIF_EXISTS(_mrt, _idx) ((_mrt)->vif_table[_idx].dev != NULL)
-#define MFC_LINES 64
struct mr_table {
struct list_head list;
@@ -69,8 +69,9 @@ struct mr_table {
struct sock __rcu *mroute_sk;
struct timer_list ipmr_expire_timer;
struct list_head mfc_unres_queue;
- struct list_head mfc_cache_array[MFC_LINES];
struct vif_device vif_table[MAXVIFS];
+ struct rhltable mfc_hash;
+ struct list_head mfc_cache_list;
int maxvif;
atomic_t cache_resolve_queue_len;
bool mroute_do_assert;
@@ -85,17 +86,48 @@ enum {
MFC_STATIC = BIT(0),
};
+struct mfc_cache_cmp_arg {
+ __be32 mfc_mcastgrp;
+ __be32 mfc_origin;
+};
+
+/**
+ * struct mfc_cache - multicast routing entries
+ * @mnode: rhashtable list
+ * @mfc_mcastgrp: destination multicast group address
+ * @mfc_origin: source address
+ * @cmparg: used for rhashtable comparisons
+ * @mfc_parent: source interface (iif)
+ * @mfc_flags: entry flags
+ * @expires: unresolved entry expire time
+ * @unresolved: unresolved cached skbs
+ * @last_assert: time of last assert
+ * @minvif: minimum VIF id
+ * @maxvif: maximum VIF id
+ * @bytes: bytes that have passed for this entry
+ * @pkt: packets that have passed for this entry
+ * @wrong_if: number of wrong source interface hits
+ * @lastuse: time of last use of the group (traffic or update)
+ * @ttls: OIF TTL threshold array
+ * @list: global entry list
+ * @rcu: used for entry destruction
+ */
struct mfc_cache {
- struct list_head list;
- __be32 mfc_mcastgrp; /* Group the entry belongs to */
- __be32 mfc_origin; /* Source of packet */
- vifi_t mfc_parent; /* Source interface */
- int mfc_flags; /* Flags on line */
+ struct rhlist_head mnode;
+ union {
+ struct {
+ __be32 mfc_mcastgrp;
+ __be32 mfc_origin;
+ };
+ struct mfc_cache_cmp_arg cmparg;
+ };
+ vifi_t mfc_parent;
+ int mfc_flags;
union {
struct {
unsigned long expires;
- struct sk_buff_head unresolved; /* Unresolved buffers */
+ struct sk_buff_head unresolved;
} unres;
struct {
unsigned long last_assert;
@@ -105,18 +137,13 @@ struct mfc_cache {
unsigned long pkt;
unsigned long wrong_if;
unsigned long lastuse;
- unsigned char ttls[MAXVIFS]; /* TTL thresholds */
+ unsigned char ttls[MAXVIFS];
} res;
} mfc_un;
+ struct list_head list;
struct rcu_head rcu;
};
-#ifdef __BIG_ENDIAN
-#define MFC_HASH(a,b) (((((__force u32)(__be32)a)>>24)^(((__force u32)(__be32)b)>>26))&(MFC_LINES-1))
-#else
-#define MFC_HASH(a,b) ((((__force u32)(__be32)a)^(((__force u32)(__be32)b)>>2))&(MFC_LINES-1))
-#endif
-
struct rtmsg;
int ipmr_get_route(struct net *net, struct sk_buff *skb,
__be32 saddr, __be32 daddr,
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 824c4fdf21eb..beacd028848c 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -299,10 +299,29 @@ static void __net_exit ipmr_rules_exit(struct net *net)
}
#endif
+static inline int ipmr_hash_cmp(struct rhashtable_compare_arg *arg,
+ const void *ptr)
+{
+ const struct mfc_cache_cmp_arg *cmparg = arg->key;
+ struct mfc_cache *c = (struct mfc_cache *)ptr;
+
+ return cmparg->mfc_mcastgrp != c->mfc_mcastgrp ||
+ cmparg->mfc_origin != c->mfc_origin;
+}
+
+static const struct rhashtable_params ipmr_rht_params = {
+ .head_offset = offsetof(struct mfc_cache, mnode),
+ .key_offset = offsetof(struct mfc_cache, cmparg),
+ .key_len = sizeof(struct mfc_cache_cmp_arg),
+ .nelem_hint = 3,
+ .locks_mul = 1,
+ .obj_cmpfn = ipmr_hash_cmp,
+ .automatic_shrinking = true,
+};
+
static struct mr_table *ipmr_new_table(struct net *net, u32 id)
{
struct mr_table *mrt;
- unsigned int i;
/* "pimreg%u" should not exceed 16 bytes (IFNAMSIZ) */
if (id != RT_TABLE_DEFAULT && id >= 1000000000)
@@ -318,10 +337,8 @@ static struct mr_table *ipmr_new_table(struct net *net, u32 id)
write_pnet(&mrt->net, net);
mrt->id = id;
- /* Forwarding cache */
- for (i = 0; i < MFC_LINES; i++)
- INIT_LIST_HEAD(&mrt->mfc_cache_array[i]);
-
+ rhltable_init(&mrt->mfc_hash, &ipmr_rht_params);
+ INIT_LIST_HEAD(&mrt->mfc_cache_list);
INIT_LIST_HEAD(&mrt->mfc_unres_queue);
setup_timer(&mrt->ipmr_expire_timer, ipmr_expire_process,
@@ -338,6 +355,7 @@ static void ipmr_free_table(struct mr_table *mrt)
{
del_timer_sync(&mrt->ipmr_expire_timer);
mroute_clean_tables(mrt, true);
+ rhltable_destroy(&mrt->mfc_hash);
kfree(mrt);
}
@@ -839,13 +857,17 @@ static struct mfc_cache *ipmr_cache_find(struct mr_table *mrt,
__be32 origin,
__be32 mcastgrp)
{
- int line = MFC_HASH(mcastgrp, origin);
+ struct mfc_cache_cmp_arg arg = {
+ .mfc_mcastgrp = mcastgrp,
+ .mfc_origin = origin
+ };
+ struct rhlist_head *tmp, *list;
struct mfc_cache *c;
- list_for_each_entry_rcu(c, &mrt->mfc_cache_array[line], list) {
- if (c->mfc_origin == origin && c->mfc_mcastgrp == mcastgrp)
- return c;
- }
+ list = rhltable_lookup(&mrt->mfc_hash, &arg, ipmr_rht_params);
+ rhl_for_each_entry_rcu(c, tmp, list, mnode)
+ return c;
+
return NULL;
}
@@ -853,13 +875,16 @@ static struct mfc_cache *ipmr_cache_find(struct mr_table *mrt,
static struct mfc_cache *ipmr_cache_find_any_parent(struct mr_table *mrt,
int vifi)
{
- int line = MFC_HASH(htonl(INADDR_ANY), htonl(INADDR_ANY));
+ struct mfc_cache_cmp_arg arg = {
+ .mfc_mcastgrp = htonl(INADDR_ANY),
+ .mfc_origin = htonl(INADDR_ANY)
+ };
+ struct rhlist_head *tmp, *list;
struct mfc_cache *c;
- list_for_each_entry_rcu(c, &mrt->mfc_cache_array[line], list)
- if (c->mfc_origin == htonl(INADDR_ANY) &&
- c->mfc_mcastgrp == htonl(INADDR_ANY) &&
- c->mfc_un.res.ttls[vifi] < 255)
+ list = rhltable_lookup(&mrt->mfc_hash, &arg, ipmr_rht_params);
+ rhl_for_each_entry_rcu(c, tmp, list, mnode)
+ if (c->mfc_un.res.ttls[vifi] < 255)
return c;
return NULL;
@@ -869,29 +894,51 @@ static struct mfc_cache *ipmr_cache_find_any_parent(struct mr_table *mrt,
static struct mfc_cache *ipmr_cache_find_any(struct mr_table *mrt,
__be32 mcastgrp, int vifi)
{
- int line = MFC_HASH(mcastgrp, htonl(INADDR_ANY));
+ struct mfc_cache_cmp_arg arg = {
+ .mfc_mcastgrp = mcastgrp,
+ .mfc_origin = htonl(INADDR_ANY)
+ };
+ struct rhlist_head *tmp, *list;
struct mfc_cache *c, *proxy;
if (mcastgrp == htonl(INADDR_ANY))
goto skip;
- list_for_each_entry_rcu(c, &mrt->mfc_cache_array[line], list)
- if (c->mfc_origin == htonl(INADDR_ANY) &&
- c->mfc_mcastgrp == mcastgrp) {
- if (c->mfc_un.res.ttls[vifi] < 255)
- return c;
-
- /* It's ok if the vifi is part of the static tree */
- proxy = ipmr_cache_find_any_parent(mrt,
- c->mfc_parent);
- if (proxy && proxy->mfc_un.res.ttls[vifi] < 255)
- return c;
- }
+ list = rhltable_lookup(&mrt->mfc_hash, &arg, ipmr_rht_params);
+ rhl_for_each_entry_rcu(c, tmp, list, mnode) {
+ if (c->mfc_un.res.ttls[vifi] < 255)
+ return c;
+
+ /* It's ok if the vifi is part of the static tree */
+ proxy = ipmr_cache_find_any_parent(mrt, c->mfc_parent);
+ if (proxy && proxy->mfc_un.res.ttls[vifi] < 255)
+ return c;
+ }
skip:
return ipmr_cache_find_any_parent(mrt, vifi);
}
+/* Look for a (S,G,iif) entry if parent != -1 */
+static struct mfc_cache *ipmr_cache_find_parent(struct mr_table *mrt,
+ __be32 origin, __be32 mcastgrp,
+ int parent)
+{
+ struct mfc_cache_cmp_arg arg = {
+ .mfc_mcastgrp = mcastgrp,
+ .mfc_origin = origin,
+ };
+ struct rhlist_head *tmp, *list;
+ struct mfc_cache *c;
+
+ list = rhltable_lookup(&mrt->mfc_hash, &arg, ipmr_rht_params);
+ rhl_for_each_entry_rcu(c, tmp, list, mnode)
+ if (parent == -1 || parent == c->mfc_parent)
+ return c;
+
+ return NULL;
+}
+
/* Allocate a multicast cache entry */
static struct mfc_cache *ipmr_cache_alloc(void)
{
@@ -1028,10 +1075,10 @@ static int ipmr_cache_report(struct mr_table *mrt,
static int ipmr_cache_unresolved(struct mr_table *mrt, vifi_t vifi,
struct sk_buff *skb)
{
+ const struct iphdr *iph = ip_hdr(skb);
+ struct mfc_cache *c;
bool found = false;
int err;
- struct mfc_cache *c;
- const struct iphdr *iph = ip_hdr(skb);
spin_lock_bh(&mfc_unres_lock);
list_for_each_entry(c, &mrt->mfc_unres_queue, list) {
@@ -1095,46 +1142,39 @@ static int ipmr_cache_unresolved(struct mr_table *mrt, vifi_t vifi,
static int ipmr_mfc_delete(struct mr_table *mrt, struct mfcctl *mfc, int parent)
{
- int line;
- struct mfc_cache *c, *next;
+ struct mfc_cache *c;
- line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
+ /* The entries are added/deleted only under RTNL */
+ rcu_read_lock();
+ c = ipmr_cache_find_parent(mrt, mfc->mfcc_origin.s_addr,
+ mfc->mfcc_mcastgrp.s_addr, parent);
+ rcu_read_unlock();
+ if (!c)
+ return -ENOENT;
+ rhltable_remove(&mrt->mfc_hash, &c->mnode, ipmr_rht_params);
+ list_del_rcu(&c->list);
+ mroute_netlink_event(mrt, c, RTM_DELROUTE);
+ ipmr_cache_free(c);
- list_for_each_entry_safe(c, next, &mrt->mfc_cache_array[line], list) {
- if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
- c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr &&
- (parent == -1 || parent == c->mfc_parent)) {
- list_del_rcu(&c->list);
- mroute_netlink_event(mrt, c, RTM_DELROUTE);
- ipmr_cache_free(c);
- return 0;
- }
- }
- return -ENOENT;
+ return 0;
}
static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
struct mfcctl *mfc, int mrtsock, int parent)
{
- bool found = false;
- int line;
struct mfc_cache *uc, *c;
+ bool found;
+ int ret;
if (mfc->mfcc_parent >= MAXVIFS)
return -ENFILE;
- line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
-
- list_for_each_entry(c, &mrt->mfc_cache_array[line], list) {
- if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
- c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr &&
- (parent == -1 || parent == c->mfc_parent)) {
- found = true;
- break;
- }
- }
-
- if (found) {
+ /* The entries are added/deleted only under RTNL */
+ rcu_read_lock();
+ c = ipmr_cache_find_parent(mrt, mfc->mfcc_origin.s_addr,
+ mfc->mfcc_mcastgrp.s_addr, parent);
+ rcu_read_unlock();
+ if (c) {
write_lock_bh(&mrt_lock);
c->mfc_parent = mfc->mfcc_parent;
ipmr_update_thresholds(mrt, c, mfc->mfcc_ttls);
@@ -1160,8 +1200,14 @@ static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
if (!mrtsock)
c->mfc_flags |= MFC_STATIC;
- list_add_rcu(&c->list, &mrt->mfc_cache_array[line]);
-
+ ret = rhltable_insert_key(&mrt->mfc_hash, &c->cmparg, &c->mnode,
+ ipmr_rht_params);
+ if (ret) {
+ pr_err("ipmr: rhtable insert error %d\n", ret);
+ ipmr_cache_free(c);
+ return ret;
+ }
+ list_add_tail_rcu(&c->list, &mrt->mfc_cache_list);
/* Check to see if we resolved a queued list. If so we
* need to send on the frames and tidy up.
*/
@@ -1191,9 +1237,9 @@ static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
/* Close the multicast socket, and clear the vif tables etc */
static void mroute_clean_tables(struct mr_table *mrt, bool all)
{
- int i;
+ struct mfc_cache *c, *tmp;
LIST_HEAD(list);
- struct mfc_cache *c, *next;
+ int i;
/* Shut down all active vif entries */
for (i = 0; i < mrt->maxvif; i++) {
@@ -1204,19 +1250,18 @@ static void mroute_clean_tables(struct mr_table *mrt, bool all)
unregister_netdevice_many(&list);
/* Wipe the cache */
- for (i = 0; i < MFC_LINES; i++) {
- list_for_each_entry_safe(c, next, &mrt->mfc_cache_array[i], list) {
- if (!all && (c->mfc_flags & MFC_STATIC))
- continue;
- list_del_rcu(&c->list);
- mroute_netlink_event(mrt, c, RTM_DELROUTE);
- ipmr_cache_free(c);
- }
+ list_for_each_entry_safe(c, tmp, &mrt->mfc_cache_list, list) {
+ if (!all && (c->mfc_flags & MFC_STATIC))
+ continue;
+ rhltable_remove(&mrt->mfc_hash, &c->mnode, ipmr_rht_params);
+ list_del_rcu(&c->list);
+ mroute_netlink_event(mrt, c, RTM_DELROUTE);
+ ipmr_cache_free(c);
}
if (atomic_read(&mrt->cache_resolve_queue_len) != 0) {
spin_lock_bh(&mfc_unres_lock);
- list_for_each_entry_safe(c, next, &mrt->mfc_unres_queue, list) {
+ list_for_each_entry_safe(c, tmp, &mrt->mfc_unres_queue, list) {
list_del(&c->list);
mroute_netlink_event(mrt, c, RTM_DELROUTE);
ipmr_destroy_unres(mrt, c);
@@ -1791,9 +1836,9 @@ static void ip_mr_forward(struct net *net, struct mr_table *mrt,
struct sk_buff *skb, struct mfc_cache *cache,
int local)
{
+ int true_vifi = ipmr_find_vif(mrt, skb->dev);
int psend = -1;
int vif, ct;
- int true_vifi = ipmr_find_vif(mrt, skb->dev);
vif = cache->mfc_parent;
cache->mfc_un.res.pkt++;
@@ -2293,34 +2338,30 @@ static int ipmr_rtm_dumproute(struct sk_buff *skb, struct netlink_callback *cb)
struct mr_table *mrt;
struct mfc_cache *mfc;
unsigned int t = 0, s_t;
- unsigned int h = 0, s_h;
unsigned int e = 0, s_e;
s_t = cb->args[0];
- s_h = cb->args[1];
- s_e = cb->args[2];
+ s_e = cb->args[1];
rcu_read_lock();
ipmr_for_each_table(mrt, net) {
if (t < s_t)
goto next_table;
- if (t > s_t)
- s_h = 0;
- for (h = s_h; h < MFC_LINES; h++) {
- list_for_each_entry_rcu(mfc, &mrt->mfc_cache_array[h], list) {
- if (e < s_e)
- goto next_entry;
- if (ipmr_fill_mroute(mrt, skb,
- NETLINK_CB(cb->skb).portid,
- cb->nlh->nlmsg_seq,
- mfc, RTM_NEWROUTE,
- NLM_F_MULTI) < 0)
- goto done;
+ list_for_each_entry_rcu(mfc, &mrt->mfc_cache_list, list) {
+ if (e < s_e)
+ goto next_entry;
+ if (ipmr_fill_mroute(mrt, skb,
+ NETLINK_CB(cb->skb).portid,
+ cb->nlh->nlmsg_seq,
+ mfc, RTM_NEWROUTE,
+ NLM_F_MULTI) < 0)
+ goto done;
next_entry:
- e++;
- }
- e = s_e = 0;
+ e++;
}
+ e = 0;
+ s_e = 0;
+
spin_lock_bh(&mfc_unres_lock);
list_for_each_entry(mfc, &mrt->mfc_unres_queue, list) {
if (e < s_e)
@@ -2337,16 +2378,15 @@ static int ipmr_rtm_dumproute(struct sk_buff *skb, struct netlink_callback *cb)
e++;
}
spin_unlock_bh(&mfc_unres_lock);
- e = s_e = 0;
- s_h = 0;
+ e = 0;
+ s_e = 0;
next_table:
t++;
}
done:
rcu_read_unlock();
- cb->args[2] = e;
- cb->args[1] = h;
+ cb->args[1] = e;
cb->args[0] = t;
return skb->len;
@@ -2590,10 +2630,8 @@ struct ipmr_mfc_iter {
struct seq_net_private p;
struct mr_table *mrt;
struct list_head *cache;
- int ct;
};
-
static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
struct ipmr_mfc_iter *it, loff_t pos)
{
@@ -2601,12 +2639,10 @@ static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
struct mfc_cache *mfc;
rcu_read_lock();
- for (it->ct = 0; it->ct < MFC_LINES; it->ct++) {
- it->cache = &mrt->mfc_cache_array[it->ct];
- list_for_each_entry_rcu(mfc, it->cache, list)
- if (pos-- == 0)
- return mfc;
- }
+ it->cache = &mrt->mfc_cache_list;
+ list_for_each_entry_rcu(mfc, &mrt->mfc_cache_list, list)
+ if (pos-- == 0)
+ return mfc;
rcu_read_unlock();
spin_lock_bh(&mfc_unres_lock);
@@ -2633,17 +2669,16 @@ static void *ipmr_mfc_seq_start(struct seq_file *seq, loff_t *pos)
it->mrt = mrt;
it->cache = NULL;
- it->ct = 0;
return *pos ? ipmr_mfc_seq_idx(net, seq->private, *pos - 1)
: SEQ_START_TOKEN;
}
static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
- struct mfc_cache *mfc = v;
struct ipmr_mfc_iter *it = seq->private;
struct net *net = seq_file_net(seq);
struct mr_table *mrt = it->mrt;
+ struct mfc_cache *mfc = v;
++*pos;
@@ -2656,19 +2691,9 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
if (it->cache == &mrt->mfc_unres_queue)
goto end_of_list;
- BUG_ON(it->cache != &mrt->mfc_cache_array[it->ct]);
-
- while (++it->ct < MFC_LINES) {
- it->cache = &mrt->mfc_cache_array[it->ct];
- if (list_empty(it->cache))
- continue;
- return list_first_entry(it->cache, struct mfc_cache, list);
- }
-
/* exhausted cache_array, show unresolved */
rcu_read_unlock();
it->cache = &mrt->mfc_unres_queue;
- it->ct = 0;
spin_lock_bh(&mfc_unres_lock);
if (!list_empty(it->cache))
@@ -2688,7 +2713,7 @@ static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
if (it->cache == &mrt->mfc_unres_queue)
spin_unlock_bh(&mfc_unres_lock);
- else if (it->cache == &mrt->mfc_cache_array[it->ct])
+ else if (it->cache == &mrt->mfc_cache_list)
rcu_read_unlock();
}
--
2.1.4
^ permalink raw reply related
* Re: [PATCH net-next 0/2] net/smc: fix typo and clc-bug
From: David Miller @ 2017-01-12 14:47 UTC (permalink / raw)
To: ubraun; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens
In-Reply-To: <20170112135715.64242-1-ubraun@linux.vnet.ibm.com>
From: Ursula Braun <ubraun@linux.vnet.ibm.com>
Date: Thu, 12 Jan 2017 14:57:13 +0100
> I received 2 bug reports for my new AF_SMC-code. Here are the fixes for them.
Series applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next v2 0/2] More flexible BPF cb access
From: David Miller @ 2017-01-12 15:00 UTC (permalink / raw)
To: daniel; +Cc: alexei.starovoitov, netdev
In-Reply-To: <cover.1484214889.git.daniel@iogearbox.net>
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Thu, 12 Jan 2017 11:51:31 +0100
> This patch improves BPF's cb access by allowing b/h/w/dw
> access variants on it. For details, please see individual
> patches.
Series applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next] lwt_bpf: bpf_lwt_prog_cmp() can be static
From: David Miller @ 2017-01-12 15:05 UTC (permalink / raw)
To: weiyj.lk; +Cc: tgraf, ast, daniel, weiyongjun1, netdev
In-Reply-To: <20170112143928.21496-1-weiyj.lk@gmail.com>
From: Wei Yongjun <weiyj.lk@gmail.com>
Date: Thu, 12 Jan 2017 14:39:28 +0000
> From: Wei Yongjun <weiyongjun1@huawei.com>
>
> Fixes the following sparse warning:
>
> net/core/lwt_bpf.c:355:5: warning:
> symbol 'bpf_lwt_prog_cmp' was not declared. Should it be static?
>
> Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Applied.
^ permalink raw reply
* Re: [PATCH RESEND net-next 00/12] s390: qeth patches
From: David Miller @ 2017-01-12 15:03 UTC (permalink / raw)
To: ubraun; +Cc: netdev, linux-s390, schwidefsky, heiko.carstens
In-Reply-To: <20170112144843.33463-1-ubraun@linux.vnet.ibm.com>
From: Ursula Braun <ubraun@linux.vnet.ibm.com>
Date: Thu, 12 Jan 2017 15:48:31 +0100
> yesterday I came up with 13 qeth patches. Since you have not been
> happy with the 13th patch, I want to make sure that at least the
> remaining 12 qeth patches can be applied to net-next. Here is the
> resend of them.
Series applied.
^ permalink raw reply
* Re: Setting link down or up in software
From: Mason @ 2017-01-12 15:16 UTC (permalink / raw)
To: netdev; +Cc: Mans Rullgard, Florian Fainelli, Andrew Lunn, Thibaud Cornic
In-Reply-To: <75bd6df2-714a-2b9b-a0e3-f5574c295377@free.fr>
On 12/01/2017 14:05, Mason wrote:
> I'm wondering what are the semantics of calling
>
> ip link set dev eth0 down
>
> I was expecting that to somehow instruct the device's ethernet driver
> to shut everything down, have the PHY tell the peer that it's going
> away, maybe even put the PHY in some low-power mode, etc.
>
> But it doesn't seem to be doing any of that on my HW.
>
> So what exactly is it supposed to do?
>
>
> And on top of that, I am seeing random occurrences of
>
> nb8800 26000.ethernet eth0: Link is Down
>
> Sometimes it is printed immediately.
> Sometimes it is printed as soon as I run "ip link set dev eth0 up" (?!)
> Sometimes it is not printed at all.
>
> I find this erratic behavior very confusing.
>
> Is it the symptom of some deeper bug?
Here's an example of "Link is Down" printed when I set link up:
At [ 62.750220] I run ip link set dev eth0 down
Then leave the system idle for 10 minutes.
At [ 646.263041] I run ip link set dev eth0 up
At [ 647.364079] it prints "Link is Down"
At [ 649.417434] it prints "Link is Up - 1Gbps/Full - flow control rx/tx"
I think whether I set up the PHY to use interrupts or polling
does have an influence on the weirdness I observe.
AFAICT, changing the interface flags is done in dev_change_flags
which calls __dev_change_flags and __dev_notify_flags
Is one of these supposed to call the device driver through a
callback at some point?
How/when is the phy_state_machine notified of the change in
interface flags?
Regards.
^ permalink raw reply
* Re: [PATCH net-next] net: ipv6: put autoconf routes into per-interface tables
From: David Ahern @ 2017-01-12 15:17 UTC (permalink / raw)
To: Lorenzo Colitti
Cc: netdev@vger.kernel.org, Maciej Żenczykowski,
Hannes Frederic Sowa, Erik Kline, YOSHIFUJI Hideaki, David Miller,
Daniel Rosenberg
In-Reply-To: <CAKD1Yr2U2Oh5yvHE7Znfc9mHzM4K6B_kU82Qctpcsmnt3y5GdQ@mail.gmail.com>
On 1/9/17 7:01 PM, Lorenzo Colitti wrote:
> On Sun, Jan 8, 2017 at 1:24 PM, David Ahern <dsa@cumulusnetworks.com> wrote:
>> Why not use the VRF capability then? create a VRF and assign the interface to it. End result is the same -- separate tables and the need to use a bind-to-device API to hit those routes.
>
> Requiring that VRFs for this creates additional complexity, because
> each network now requires its own VRF. That means that the connection
> manager must create the VRF before the interface comes up and receives
> the RA.
>
> In some cases this might not be possible. For example, consider a tun
> interface that's created by a different process such as a VPN client.
> In this case the connection manager doesn't know the interface name,
> and the VPN client doesn't know to create the VRF, so if the tun
> interface gets an RA after the tun is created but
Have you looked at adding basic l3mdev capabilities to tun? in this case just l3mdev_fib_table needs be implemented. On interface create push down a table id and set the IFF_L3MDEV_MASTER flag.
^ permalink raw reply
* Re: [PATCH] [net] net/mlx5e: fix another -Wmaybe-uninitialized warning
From: Or Gerlitz @ 2017-01-12 15:21 UTC (permalink / raw)
To: Arnd Bergmann
Cc: Saeed Mahameed, Hadar Hen Zion, David S . Miller, netdev,
linux-kernel
In-Reply-To: <20170111211451.2705705-1-arnd@arndb.de>
On 1/11/2017 11:14 PM, Arnd Bergmann wrote:
> As found by Olof's build bot, today's mainline kernel gained a harmless
> warning about a potential uninitalied variable reference:
>
> drivers/net/ethernet/mellanox/mlx5/core/en_tc.c: In function 'parse_tc_fdb_actions':
> drivers/net/ethernet/mellanox/mlx5/core/en_tc.c:769:13: warning: 'out_dev' may be used uninitialized in this function [-Wmaybe-uninitialized]
> drivers/net/ethernet/mellanox/mlx5/core/en_tc.c:811:21: note: 'out_dev' was declared here
>
> This was introduced through the addition of an 'IS_ERR/PTR_ERR' pair that
> gcc is unfortunately unable to completely figure out. Replacing it with
> PTR_ERR_OR_ZERO makes the code more understandable to gcc so it no longer
> warns.
can you elaborate on this a little further?
> Hadar Hen Zion already attempted to fix the warning earlier by adding
> fake initializations, but that ended up just making the code worse without
> fully addressing all warnings, so I'm reverting it now that it is no longer needed.
ok, so if your approach eliminates the warning on out_dev and also on
the variables for which Hadar added the faked initializers, I guess we
should be fine with this change (saw your reply on my other comment),
just another question:
> In order to avoid pulling a variable declaration into the #ifdef, I'm
> removing it in favor of a more readable 'if()' statement here that has the same effect.
When I build here without CONFIG_INET in my system, the build goes fine
with this approach. However, we're pretty sure that in the past we got
0-day report from the kbuild test robot where he was unhappy that we
make the ip_route_output_key call without being wrapped with that #if
IS_ENABLED(CONFIG_INET) -- so, we don't want to go there again... thoughts?
Or.
^ permalink raw reply
* Correct method for initializing Pause and Asymmetrical Pause support in phy drivers
From: Marc Bertola @ 2017-01-12 15:21 UTC (permalink / raw)
To: netdev
Hello netdev list,
I am currently investigating a problem related to Ethernet
auto-negotiation of Pause and Asymmetrical Pause capabilities.
TL;DR: I am using a Picozed system-on-module with a Xilinx Gigabit
Ethernet MAC and a Marvell PHY. It does not appear to be advertising
support for Pause and Asym Pause, which seems strange to me given that
this is relatively recent hardware. I suspect that may be due to a
problem in the way phydev->supported is initialized in
drivers/net/phy/marvell.c.
I am trying to confirm what the proper method is to initialize
phydev->supported such that it advertises SUPPORTED_Pause and
SUPPORTED_Asym_Pause. Adding these flags to (phy_driver).features
seems to work, but I would like to confirm with people who are more
knowledgeable than me in this regard.
Read on for details about what I have observed and tried so far.
# The System #
The application I am working on uses Avnet's Picozed 7020
System-on-Module (SOM), which contains:
* An on-chip MAC (on a Xilinx Zynq 7000 chip)
* A Marvell 1512 Alaska PHY.
* A daughtercard that provides the actual RJ45 connector.
The Zynq is running a Xilinx fork of Linux. I am working with the
following drivers:
The MAC is the built-in Gigabit Ethernet MAC on a Xilinx Zynq 7000
chip. It uses the xemacps.c driver, which can be found on Xilinx's
official Linux fork:
* RAW: https://raw.githubusercontent.com/Xilinx/linux-xlnx/master/drivers/net/ethernet/xilinx/xilinx_emacps.c
* GITHUB: https://github.com/Xilinx/linux-xlnx/blob/master/drivers/net/ethernet/xilinx/xilinx_emacps.c
The PHY is a Marvell 1512 Alaska device that comes on the Picozed 7020
SOM that is the heart of the application I am working on.
The version of the driver I am using for this PHY can be found here
(note that this version is slightly different (older?) to the mainline
Linux repo):
* RAW: https://raw.githubusercontent.com/Xilinx/linux-xlnx/master/drivers/net/phy/marvell.c
* GITHUB: https://github.com/Xilinx/linux-xlnx/blob/master/drivers/net/phy/marvell.c
# The Problem: No Flow Control via Auto-Negotiation #
I have noticed that when I connect to the Picozed using a PC, the
connection speed and duplex are negotiated correctly, and the signal
integrity is good. However, its autonegotiation capabilities do not
report support for the Pause or Asymmetrical Pause capabilities. Flow
control is thus disabled as a result. I verified this by dumping the
Link Partner capabilities PHY register on my PC.
If I connect another regular PC or a smart switch (on which I have
enabled flow control) to my PC, flow control capability is reported
and thus Pause frames are enabled.
Based on the Zynq's Technical Reference Manual, the MAC supports Pause
frames and Asymmetrical Pause, so I am working with the assumption
that these features should be advertised, contrary to what I am
seeing.
I spent most of the day looking at the PHY abstraction layer, the
marvell driver, and the xemacps driver to figure out where the missing
flow control capability information needed to be added. I also looked
at the phy.txt where I found the following:
"Now just make sure that phydev->supported and phydev->advertising
have any values pruned from them which don't make sense for your
controller a 10/100 controller may be connected to a gigabit capable
PHY, so you would need to mask off SUPPORTED_1000baseT*). See
include/linux/ethtool.h for definitions for these bitfields. Note that
you should not SET any bits, or the PHY may get put into an
unsupported state."
Source: https://raw.githubusercontent.com/Xilinx/linux-xlnx/master/Documentation/networking/phy.txt
So my understanding is as follows:
1. The PHY driver sets all of the flags for the capabilities it
supports in phydev->supported.
2. The MAC driver then prunes the capabilities it does not support
from phydev->supported to see what can be safely advertised.
In xemacps.c, the following code appears to be performing this
pruning, by removing all capabilities other than PHY_GBIT_FEATURES and
the Flow Control capability bits.
phydev->supported &= (PHY_GBIT_FEATURES | SUPPORTED_Pause |
SUPPORTED_Asym_Pause);
So, if the PHY had advertised that it supported Flow Control / Pause
Frames, these capabilities would have been preserved. However, with
some variable dumping to dmesg, I can see that SUPPORTED_Pause and
SUPPORTED_Asym_Pause are not present in phydev->supported. What I
observe is that phydev->supported has a value of 0x02ff, and we would
expect bits 13 and 14 to be set, resulting in 0x62ff.
I dug a bit deeper to see how the PHY driver populates the value of
phydev->supported before it gets passed to the MAC for pruning. I
found that it comes from the .features field in the phy_driver structs
defined at the bottom of the marvell.c file. In the version I am
using, this field only contains PHY_GBIT_FEATURES, which explains why
the SUPPORTED_Pause and SUPPORTED_Asym_Pause flags are not set.
{
.phy_id = MARVELL_PHY_ID_88E1510,
.phy_id_mask = MARVELL_PHY_ID_MASK,
.name = "Marvell 88E1510",
.features = PHY_GBIT_FEATURES, <--- This!
.flags = PHY_HAS_INTERRUPT,
...
}
# Solution Ideas... #
My first question is: Was this done on purpose? I find it hard to
believe that a relatively recent PHY would not support Pause and Asym
Pause advertisement. If this is indeed an accidental oversight, what
would the proper fix be?
I have tried to OR in the appropriate SUPPORTED_Pause and
SUPPORTED_Asym_Pause values into the .features field, like this:
{
...
.features = (PHY_GBIT_FEATURES | SUPPORTED_Pause | SUPPORTED_Asym_Pause),
...
}
Now I see that auto-negotiation works as I would expect it to: Flow
Control is advertised in the Link Partner register, and my PC reports
that flow control is enabled. One caveat is that ethtool -a eth0 on
the Zynq still claims flow control is off, so I might still have some
adjustments to make for this to be displayed correctly.
What are your thoughts on this? Is my reasoning correct, or is there a
different best practice to follow in this situation?
This is my first post on this mailing list. Thanks in advance for your
help and patience.
Marc Bertola, Prolucid Technologies
^ permalink raw reply
* Re: [PATCH net] ravb: Remove Rx overflow log messages
From: David Miller @ 2017-01-12 15:22 UTC (permalink / raw)
To: horms+renesas; +Cc: sergei.shtylyov, magnus.damm, netdev, linux-renesas-soc
In-Reply-To: <20170112122106.GA32727@verge.net.au>
From: Simon Horman <horms+renesas@verge.net.au>
Date: Thu, 12 Jan 2017 13:21:06 +0100
> From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
>
> Remove Rx overflow log messages as in an environment where logging results
> in network traffic logging may cause further overflows.
>
> Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
> Signed-off-by: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> [simon: reworked changelog]
> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
> Acked-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH v2 2/2] stmmac: rename it to synopsys
From: David Miller @ 2017-01-12 15:26 UTC (permalink / raw)
To: Joao.Pinto
Cc: alexandre.torgue, f.fainelli, lars.persson, niklass,
peppe.cavallaro, netdev
In-Reply-To: <dbb59d07-4ab4-37d2-5e70-073091749445@synopsys.com>
I don't understand at all why it is so important to change the name of
these files nor the directory they live in.
What bonafide benefit will users receive if we do this?
The only clear part is the downside, which is that it is going to make
it painful to browse source history and backport bug fixes.
Please, let's not do this.
Thanks.
^ permalink raw reply
* [PATCH] net: thunderx: acpi: fix LMAC initialization
From: Vadim Lomovtsev @ 2017-01-12 15:28 UTC (permalink / raw)
To: linux-arm-kernel, netdev, linux-kernel, sgoutham, rric,
Radha.Chintakuntla
Cc: Vadim Lomovtsev
While probing BGX we requesting appropriate QLM for it's configuration
and get LMAC count by that request. Then, while reading configured
MAC values from SSDT table we need to save them in proper mapping:
BGX[i]->lmac[j].mac = <MAC value>
to later provide for initialization stuff. In order to fill
such mapping properly we need to add lmac index to be used while
acpi initialization since at this moment bgx->lmac_count already contains
actual value.
Signed-off-by: Vadim Lomovtsev <Vadim.Lomovtsev@caviumnetworks.com>
---
drivers/net/ethernet/cavium/thunder/thunder_bgx.c | 11 ++++++-----
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
index be30ad0..a3f4f83 100644
--- a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
+++ b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
@@ -47,8 +47,9 @@ struct lmac {
struct bgx {
u8 bgx_id;
struct lmac lmac[MAX_LMAC_PER_BGX];
- int lmac_count;
+ u8 lmac_count;
u8 max_lmac;
+ u8 acpi_lmac_idx;
void __iomem *reg_base;
struct pci_dev *pdev;
bool is_dlm;
@@ -1073,13 +1074,13 @@ static acpi_status bgx_acpi_register_phy(acpi_handle handle,
if (acpi_bus_get_device(handle, &adev))
goto out;
- acpi_get_mac_address(dev, adev, bgx->lmac[bgx->lmac_count].mac);
+ acpi_get_mac_address(dev, adev, bgx->lmac[bgx->acpi_lmac_idx].mac);
- SET_NETDEV_DEV(&bgx->lmac[bgx->lmac_count].netdev, dev);
+ SET_NETDEV_DEV(&bgx->lmac[bgx->acpi_lmac_idx].netdev, dev);
- bgx->lmac[bgx->lmac_count].lmacid = bgx->lmac_count;
+ bgx->lmac[bgx->acpi_lmac_idx].lmacid = bgx->acpi_lmac_idx;
+ bgx->acpi_lmac_idx++; /* move to next LMAC */
out:
- bgx->lmac_count++;
return AE_OK;
}
--
1.8.3.1
^ permalink raw reply related
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