* [PATCH net-next 1/7] timecounter: provide a macro to initialize the cyclecounter mask field.
From: Richard Cochran @ 2015-01-01 10:39 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, David Miller, Jeff Kirsher, John Stultz,
Thomas Gleixner
In-Reply-To: <cover.1420108214.git.richardcochran@gmail.com>
There is no need for users of the timecounter/cyclecounter code to include
clocksource.h just for a single macro.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
---
include/linux/timecounter.h | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/include/linux/timecounter.h b/include/linux/timecounter.h
index 74f4549..4382035 100644
--- a/include/linux/timecounter.h
+++ b/include/linux/timecounter.h
@@ -19,6 +19,9 @@
#include <linux/types.h>
+/* simplify initialization of mask field */
+#define CYCLECOUNTER_MASK(bits) (cycle_t)((bits) < 64 ? ((1ULL<<(bits))-1) : -1)
+
/**
* struct cyclecounter - hardware abstraction for a free running counter
* Provides completely state-free accessors to the underlying hardware.
@@ -29,7 +32,7 @@
* @read: returns the current cycle value
* @mask: bitmask for two's complement
* subtraction of non 64 bit counters,
- * see CLOCKSOURCE_MASK() helper macro
+ * see CYCLECOUNTER_MASK() helper macro
* @mult: cycle to nanosecond multiplier
* @shift: cycle to nanosecond divisor (power of two)
*/
--
1.7.10.4
^ permalink raw reply related
* [PATCH net-next 0/7] Fixing the "Time Counter fixes and improvements"
From: Richard Cochran @ 2015-01-01 10:39 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, David Miller, Jeff Kirsher, John Stultz,
Thomas Gleixner
In-Reply-To: <20141231.183347.862533634176009078.davem@davemloft.net>
Dave,
I did not catch the missing includes in my x86 and arm testing,
because those archs somehow do include clocksource.h for the drivers
in question. Sorry.
This is how I would like to fix the header fallout. We really should
decouple the timecounter/cyclecounter code from the clocksource code
where possible.
Thanks,
Richard
Richard Cochran (7):
timecounter: provide a macro to initialize the cyclecounter mask
field.
bnx2x: convert to CYCLECOUNTER_MASK macro.
e1000e: convert to CYCLECOUNTER_MASK macro.
igb: convert to CYCLECOUNTER_MASK macro.
ixgbe: convert to CYCLECOUNTER_MASK macro.
mlx4: include clocksource.h again
microblaze: include the new timecounter header.
arch/microblaze/kernel/timer.c | 1 +
drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c | 2 +-
drivers/net/ethernet/intel/e1000e/netdev.c | 2 +-
drivers/net/ethernet/intel/igb/igb_ptp.c | 4 ++--
drivers/net/ethernet/intel/ixgbe/ixgbe_ptp.c | 2 +-
drivers/net/ethernet/mellanox/mlx4/en_clock.c | 1 +
include/linux/timecounter.h | 5 ++++-
7 files changed, 11 insertions(+), 6 deletions(-)
--
1.7.10.4
^ permalink raw reply
* [PULL] vhost: cleanups and fixes
From: Michael S. Tsirkin @ 2015-01-01 12:26 UTC (permalink / raw)
To: Linus Torvalds; +Cc: kvm, virtualization, netdev, linux-kernel, mst, rusty
The following changes since commit b7392d2247cfe6771f95d256374f1a8e6a6f48d6:
Linux 3.19-rc2 (2014-12-28 16:49:37 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost.git tags/for_linus
for you to fetch changes up to 5d9a07b0de512b77bf28d2401e5fe3351f00a240:
vhost: relax used address alignment (2014-12-29 10:55:06 +0200)
----------------------------------------------------------------
vhost: virtio 1.0 bugfix
There's a single change here, fixing a vhost bug where vhost initialization
fails due to used ring alignment check being too strict.
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
----------------------------------------------------------------
Michael S. Tsirkin (2):
virtio_ring: document alignment requirements
vhost: relax used address alignment
include/uapi/linux/virtio_ring.h | 7 +++++++
drivers/vhost/vhost.c | 10 +++++++---
2 files changed, 14 insertions(+), 3 deletions(-)
^ permalink raw reply
* Re: [PATCH] TCP: Add support for TCP Stealth
From: Daniel Borkmann @ 2015-01-01 15:25 UTC (permalink / raw)
To: Julian Kirsch
Cc: netdev, Christian Grothoff, Jacob Appelbaum, Pavel Emelyanov
In-Reply-To: <54A470B3.3010501@sec.in.tum.de>
Hi Julian,
On 12/31/2014 10:54 PM, Julian Kirsch wrote:
...
> one year ago [0] we tried to convince you to add support for a new
> socket option to the linux kernel. Equipped with an improved version of
> our patch we're back to accomplish this task today. :-)
>
> TCP Stealth is a modern variant of port knocking which borrows
> techniques from network steganography to enable clients to authenticate
> themselves towards a server on TCP level. You can find technical details
> in an rfc draft we wrote earlier this year [1] and in my master's thesis
> [2]. In summary, TCP Stealth derives authentication information from a
> pre-shared secret and embeds it into the ISN sent along with the first
> SYN from the client.
/me wondering (haven't tried that though) ... have you considered f.e.
building a library using a raw packet socket with a BPF filter to capture
SYN packets and then TCP_REPAIR [1] to build a full-blown TCP socket out
of it in case of a correct authentication from the ISN?
Thanks,
Daniel
[1] http://www.criu.org/TCP_connection
> Our motivation is simple: During this year we gained hard evidence on
> secret services actively port scanning the internets followed by
> exploitation of your services using 0-day exploits [3, 4]. We don't want
> our machines to be turned into relays from where they continue to
> cascade their attacks. TCP Stealth makes port scanning more expensive by
> a factor of 2^31 (on average).
>
> A copy of this patch as well as patches for several user space
> applications can be found on the project's home page [5].
>
> All the best for the upcoming year,
> Julian & Christian
>
>
>
> [0] https://lkml.org/lkml/2013/12/10/1155
> [1] https://datatracker.ietf.org/doc/draft-kirsch-ietf-tcp-stealth/
> [2] https://gnunet.org/kirsch2014knock
> [3]
> http://www.heise.de/ct/artikel/NSA-GCHQ-The-HACIENDA-Program-for-Internet-Colonization-2292681.html
> [4]
> https://firstlook.org/theintercept/2014/12/13/belgacom-hack-gchq-inside-story/
> [5] https://gnunet.org/knock
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1
>
> iQEcBAEBAgAGBQJUpHCvAAoJENwkOWttRRA4g10IALbJZU9/5Gp8tVdpXqbkOIMp
> Kz+yOMyYULqYeM8yguSBZjZLbaz/VAS7SNpQxKGU+W0aAXa22FsSfVoUU7wqp3NT
> 3EGRuPkMaJkQ66IP8MtX+6/hSeWSh78tEaIFWVjyutihPyQGz0LefFc66gm54X4T
> s8IYW7jKFhNmmROu9CXLTxq4B5t2v+Evv/qWqotZqR1t3IbIUmZAiKrlkMRd7dtM
> SaS5JwFeiObxn+0M/7javQCAhfgPXYEOU0QKAGY55MXcPAner/5PuExIZdOJ41R3
> XD9tgoLGhHEiQkxj0/bP2cs3Cl5xfJl9t2iecVfTIR7PytaTJ/kFuE4gNgWEcTA=
> =T6/C
> -----END PGP SIGNATURE-----
>
^ permalink raw reply
* [PATCH] net: wireless: ipw2x00: ipw2100.c: Remove some unused functions
From: Rickard Strandqvist @ 2015-01-01 15:32 UTC (permalink / raw)
To: Stanislav Yakovlev, Kalle Valo
Cc: Rickard Strandqvist, linux-wireless, netdev, linux-kernel
Removes some functions that are not used anywhere:
write_nic_dword_auto_inc() write_nic_auto_inc_address()
This was partially found by using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
---
drivers/net/wireless/ipw2x00/ipw2100.c | 11 -----------
1 file changed, 11 deletions(-)
diff --git a/drivers/net/wireless/ipw2x00/ipw2100.c b/drivers/net/wireless/ipw2x00/ipw2100.c
index 6fabea0..fa88839 100644
--- a/drivers/net/wireless/ipw2x00/ipw2100.c
+++ b/drivers/net/wireless/ipw2x00/ipw2100.c
@@ -433,17 +433,6 @@ static inline void write_nic_byte(struct net_device *dev, u32 addr, u8 val)
write_register_byte(dev, IPW_REG_INDIRECT_ACCESS_DATA, val);
}
-static inline void write_nic_auto_inc_address(struct net_device *dev, u32 addr)
-{
- write_register(dev, IPW_REG_AUTOINCREMENT_ADDRESS,
- addr & IPW_REG_INDIRECT_ADDR_MASK);
-}
-
-static inline void write_nic_dword_auto_inc(struct net_device *dev, u32 val)
-{
- write_register(dev, IPW_REG_AUTOINCREMENT_DATA, val);
-}
-
static void write_nic_memory(struct net_device *dev, u32 addr, u32 len,
const u8 * buf)
{
--
1.7.10.4
^ permalink raw reply related
* Re: [PATCH] TCP: Add support for TCP Stealth
From: Christian Grothoff @ 2015-01-01 15:32 UTC (permalink / raw)
To: Daniel Borkmann, Julian Kirsch; +Cc: netdev, Jacob Appelbaum, Pavel Emelyanov
In-Reply-To: <54A566F2.4070401@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 783 bytes --]
Dear Daniel,
That approach is highly vulnerable to timing attacks, and doesn't answer
how TCP clients without special capabilities could set the ISN correctly
either. Playing with raw sockets is the kind of geeky hack that is
unlikely to give us the combination of usability and security required
to significantly reduce the ongoing large-scale compromise of network
equipment by spy agencies.
Christian
On 01/01/2015 04:25 PM, Daniel Borkmann wrote:
>
> /me wondering (haven't tried that though) ... have you considered f.e.
> building a library using a raw packet socket with a BPF filter to capture
> SYN packets and then TCP_REPAIR [1] to build a full-blown TCP socket out
> of it in case of a correct authentication from the ISN?
>
> Thanks,
> Daniel
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply
* [PATCH] net: wireless: b43legacy: radio.c: Remove unused function
From: Rickard Strandqvist @ 2015-01-01 15:46 UTC (permalink / raw)
To: Larry Finger, Stefano Brivio
Cc: Rickard Strandqvist, Kalle Valo, linux-wireless, b43-dev, netdev,
linux-kernel
Remove the function b43legacy_radio_set_tx_iq() that is not used anywhere.
This was partially found by using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
---
drivers/net/wireless/b43legacy/radio.c | 19 -------------------
drivers/net/wireless/b43legacy/radio.h | 1 -
2 files changed, 20 deletions(-)
diff --git a/drivers/net/wireless/b43legacy/radio.c b/drivers/net/wireless/b43legacy/radio.c
index 8961776..9501420 100644
--- a/drivers/net/wireless/b43legacy/radio.c
+++ b/drivers/net/wireless/b43legacy/radio.c
@@ -1743,25 +1743,6 @@ u16 freq_r3A_value(u16 frequency)
return value;
}
-void b43legacy_radio_set_tx_iq(struct b43legacy_wldev *dev)
-{
- static const u8 data_high[5] = { 0x00, 0x40, 0x80, 0x90, 0xD0 };
- static const u8 data_low[5] = { 0x00, 0x01, 0x05, 0x06, 0x0A };
- u16 tmp = b43legacy_radio_read16(dev, 0x001E);
- int i;
- int j;
-
- for (i = 0; i < 5; i++) {
- for (j = 0; j < 5; j++) {
- if (tmp == (data_high[i] | data_low[j])) {
- b43legacy_phy_write(dev, 0x0069, (i - j) << 8 |
- 0x00C0);
- return;
- }
- }
- }
-}
-
int b43legacy_radio_selectchannel(struct b43legacy_wldev *dev,
u8 channel,
int synthetic_pu_workaround)
diff --git a/drivers/net/wireless/b43legacy/radio.h b/drivers/net/wireless/b43legacy/radio.h
index bccb3d7..dd2976d 100644
--- a/drivers/net/wireless/b43legacy/radio.h
+++ b/drivers/net/wireless/b43legacy/radio.h
@@ -92,7 +92,6 @@ void b43legacy_nrssi_hw_write(struct b43legacy_wldev *dev, u16 offset, s16 val);
void b43legacy_nrssi_hw_update(struct b43legacy_wldev *dev, u16 val);
void b43legacy_nrssi_mem_update(struct b43legacy_wldev *dev);
-void b43legacy_radio_set_tx_iq(struct b43legacy_wldev *dev);
u16 b43legacy_radio_calibrationvalue(struct b43legacy_wldev *dev);
#endif /* B43legacy_RADIO_H_ */
--
1.7.10.4
^ permalink raw reply related
* Re: [PATCH net-next 0/7] Fixing the "Time Counter fixes and improvements"
From: Sedat Dilek @ 2015-01-01 16:30 UTC (permalink / raw)
To: Richard Cochran; +Cc: David Miller, netdev@vger.kernel.org
Richard wrote:
Dave,
I did not catch the missing includes in my x86 and arm testing,
because those archs somehow do include clocksource.h for the drivers
in question. Sorry.
This is how I would like to fix the header fallout. We really should
decouple the timecounter/cyclecounter code from the clocksource code
where possible.
Thanks,
Richard
Richard Cochran (7):
timecounter: provide a macro to initialize the cyclecounter mask
field.
bnx2x: convert to CYCLECOUNTER_MASK macro.
e1000e: convert to CYCLECOUNTER_MASK macro.
igb: convert to CYCLECOUNTER_MASK macro.
ixgbe: convert to CYCLECOUNTER_MASK macro.
mlx4: include clocksource.h again
microblaze: include the new timecounter header.
[...]
With this conversion those 2 commits in net-next.git#master are obsolete now...
e1000e: Include clocksource.h to get CLOCKSOURCE_MASK.
igb_ptp: Include clocksource.h to get CLOCKSOURCE_MASK.
- Sedat -
^ permalink raw reply
* [PATCH] net: ethernet: chelsio: cxgb3: mc5.c: Remove some unused functions
From: Rickard Strandqvist @ 2015-01-01 16:49 UTC (permalink / raw)
To: Santosh Raspatur, netdev; +Cc: Rickard Strandqvist, linux-kernel
Removes some functions that are not used anywhere:
dbgi_rd_rsp3() dbgi_wr_addr3()
This was partially found by using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
---
drivers/net/ethernet/chelsio/cxgb3/mc5.c | 16 ----------------
1 file changed, 16 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb3/mc5.c b/drivers/net/ethernet/chelsio/cxgb3/mc5.c
index e13b7fe..338301b 100644
--- a/drivers/net/ethernet/chelsio/cxgb3/mc5.c
+++ b/drivers/net/ethernet/chelsio/cxgb3/mc5.c
@@ -97,14 +97,6 @@ static int mc5_cmd_write(struct adapter *adapter, u32 cmd)
F_DBGIRSPVALID, 1, MAX_WRITE_ATTEMPTS, 1);
}
-static inline void dbgi_wr_addr3(struct adapter *adapter, u32 v1, u32 v2,
- u32 v3)
-{
- t3_write_reg(adapter, A_MC5_DB_DBGI_REQ_ADDR0, v1);
- t3_write_reg(adapter, A_MC5_DB_DBGI_REQ_ADDR1, v2);
- t3_write_reg(adapter, A_MC5_DB_DBGI_REQ_ADDR2, v3);
-}
-
static inline void dbgi_wr_data3(struct adapter *adapter, u32 v1, u32 v2,
u32 v3)
{
@@ -113,14 +105,6 @@ static inline void dbgi_wr_data3(struct adapter *adapter, u32 v1, u32 v2,
t3_write_reg(adapter, A_MC5_DB_DBGI_REQ_DATA2, v3);
}
-static inline void dbgi_rd_rsp3(struct adapter *adapter, u32 *v1, u32 *v2,
- u32 *v3)
-{
- *v1 = t3_read_reg(adapter, A_MC5_DB_DBGI_RSP_DATA0);
- *v2 = t3_read_reg(adapter, A_MC5_DB_DBGI_RSP_DATA1);
- *v3 = t3_read_reg(adapter, A_MC5_DB_DBGI_RSP_DATA2);
-}
-
/*
* Write data to the TCAM register at address (0, 0, addr_lo) using the TCAM
* command cmd. The data to be written must have been set up by the caller.
--
1.7.10.4
^ permalink raw reply related
* [PATCH] net: ethernet: micrel: ksz884x.c: Remove some unused functions
From: Rickard Strandqvist @ 2015-01-01 17:00 UTC (permalink / raw)
To: David S. Miller, Andrew Morton
Cc: Rickard Strandqvist, Wilfried Klaebe, Eric W. Biederman,
Benoit Taine, Julia Lawall, Manuel Schölling, Joe Perches,
netdev, linux-kernel
Removes some functions that are not used anywhere:
hw_w_phy_link_md() hw_r_phy_link_md() hw_w_phy_polarity()
hw_r_phy_polarity() hw_w_phy_crossover() hw_r_phy_crossover()
hw_r_phy_rem_cap() hw_w_phy_auto_neg() hw_r_phy_auto_neg()
hw_r_phy_link_stat() sw_get_addr() port_chk_prio() port_chk_replace_vid()
port_chk_802_1p() port_chk_diffserv() sw_chk_unk_def_port()
sw_cfg_unk_def_port() sw_cfg_chk_unk_def_deliver() sw_cfg_unk_def_deliver()
port_chk_in_filter() port_chk_dis_non_vid() port_cfg_in_filter()
port_cfg_dis_non_vid() port_chk_rmv_tag() port_chk_ins_tag()
port_cfg_rmv_tag() port_cfg_ins_tag() sw_flush_dyn_mac_table() port_cfg_tx()
port_cfg_rx() port_chk_force_flow_ctrl() port_chk_back_pressure()
port_cfg_force_flow_ctrl() port_chk_broad_storm() hw_ena_intr_bit()
This was partially found by using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
---
drivers/net/ethernet/micrel/ksz884x.c | 217 ---------------------------------
1 file changed, 217 deletions(-)
diff --git a/drivers/net/ethernet/micrel/ksz884x.c b/drivers/net/ethernet/micrel/ksz884x.c
index f1ebed6c..582973f 100644
--- a/drivers/net/ethernet/micrel/ksz884x.c
+++ b/drivers/net/ethernet/micrel/ksz884x.c
@@ -1551,15 +1551,6 @@ static void hw_turn_on_intr(struct ksz_hw *hw, u32 bit)
hw_set_intr(hw, hw->intr_mask);
}
-static inline void hw_ena_intr_bit(struct ksz_hw *hw, uint interrupt)
-{
- u32 read_intr;
-
- read_intr = readl(hw->io + KS884X_INTERRUPTS_ENABLE);
- hw->intr_set = read_intr | interrupt;
- writel(hw->intr_set, hw->io + KS884X_INTERRUPTS_ENABLE);
-}
-
static inline void hw_read_intr(struct ksz_hw *hw, uint *status)
{
*status = readl(hw->io + KS884X_INTERRUPTS_STATUS);
@@ -2125,12 +2116,6 @@ static inline void port_cfg_broad_storm(struct ksz_hw *hw, int p, int set)
KS8842_PORT_CTRL_1_OFFSET, PORT_BROADCAST_STORM, set);
}
-static inline int port_chk_broad_storm(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_BROADCAST_STORM);
-}
-
/* Driver set switch broadcast storm protection at 10% rate. */
#define BROADCAST_STORM_PROTECTION_RATE 10
@@ -2283,24 +2268,6 @@ static inline void port_cfg_back_pressure(struct ksz_hw *hw, int p, int set)
KS8842_PORT_CTRL_2_OFFSET, PORT_BACK_PRESSURE, set);
}
-static inline void port_cfg_force_flow_ctrl(struct ksz_hw *hw, int p, int set)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_FORCE_FLOW_CTRL, set);
-}
-
-static inline int port_chk_back_pressure(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_BACK_PRESSURE);
-}
-
-static inline int port_chk_force_flow_ctrl(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_FORCE_FLOW_CTRL);
-}
-
/* Spanning Tree */
static inline void port_cfg_dis_learn(struct ksz_hw *hw, int p, int set)
@@ -2309,82 +2276,11 @@ static inline void port_cfg_dis_learn(struct ksz_hw *hw, int p, int set)
KS8842_PORT_CTRL_2_OFFSET, PORT_LEARN_DISABLE, set);
}
-static inline void port_cfg_rx(struct ksz_hw *hw, int p, int set)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_RX_ENABLE, set);
-}
-
-static inline void port_cfg_tx(struct ksz_hw *hw, int p, int set)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_TX_ENABLE, set);
-}
-
static inline void sw_cfg_fast_aging(struct ksz_hw *hw, int set)
{
sw_cfg(hw, KS8842_SWITCH_CTRL_1_OFFSET, SWITCH_FAST_AGING, set);
}
-static inline void sw_flush_dyn_mac_table(struct ksz_hw *hw)
-{
- if (!(hw->overrides & FAST_AGING)) {
- sw_cfg_fast_aging(hw, 1);
- mdelay(1);
- sw_cfg_fast_aging(hw, 0);
- }
-}
-
-/* VLAN */
-
-static inline void port_cfg_ins_tag(struct ksz_hw *hw, int p, int insert)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_INSERT_TAG, insert);
-}
-
-static inline void port_cfg_rmv_tag(struct ksz_hw *hw, int p, int remove)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_REMOVE_TAG, remove);
-}
-
-static inline int port_chk_ins_tag(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_INSERT_TAG);
-}
-
-static inline int port_chk_rmv_tag(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_REMOVE_TAG);
-}
-
-static inline void port_cfg_dis_non_vid(struct ksz_hw *hw, int p, int set)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_DISCARD_NON_VID, set);
-}
-
-static inline void port_cfg_in_filter(struct ksz_hw *hw, int p, int set)
-{
- port_cfg(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_INGRESS_VLAN_FILTER, set);
-}
-
-static inline int port_chk_dis_non_vid(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_DISCARD_NON_VID);
-}
-
-static inline int port_chk_in_filter(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_INGRESS_VLAN_FILTER);
-}
-
/* Mirroring */
static inline void port_cfg_mirror_sniffer(struct ksz_hw *hw, int p, int set)
@@ -2422,28 +2318,6 @@ static void sw_init_mirror(struct ksz_hw *hw)
sw_cfg_mirror_rx_tx(hw, 0);
}
-static inline void sw_cfg_unk_def_deliver(struct ksz_hw *hw, int set)
-{
- sw_cfg(hw, KS8842_SWITCH_CTRL_7_OFFSET,
- SWITCH_UNK_DEF_PORT_ENABLE, set);
-}
-
-static inline int sw_cfg_chk_unk_def_deliver(struct ksz_hw *hw)
-{
- return sw_chk(hw, KS8842_SWITCH_CTRL_7_OFFSET,
- SWITCH_UNK_DEF_PORT_ENABLE);
-}
-
-static inline void sw_cfg_unk_def_port(struct ksz_hw *hw, int port, int set)
-{
- port_cfg_shift(hw, port, KS8842_SWITCH_CTRL_7_OFFSET, 0, set);
-}
-
-static inline int sw_chk_unk_def_port(struct ksz_hw *hw, int port)
-{
- return port_chk_shift(hw, port, KS8842_SWITCH_CTRL_7_OFFSET, 0);
-}
-
/* Priority */
static inline void port_cfg_diffserv(struct ksz_hw *hw, int p, int set)
@@ -2470,30 +2344,6 @@ static inline void port_cfg_prio(struct ksz_hw *hw, int p, int set)
KS8842_PORT_CTRL_1_OFFSET, PORT_PRIO_QUEUE_ENABLE, set);
}
-static inline int port_chk_diffserv(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_DIFFSERV_ENABLE);
-}
-
-static inline int port_chk_802_1p(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_802_1P_ENABLE);
-}
-
-static inline int port_chk_replace_vid(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_2_OFFSET, PORT_USER_PRIORITY_CEILING);
-}
-
-static inline int port_chk_prio(struct ksz_hw *hw, int p)
-{
- return port_chk(hw, p,
- KS8842_PORT_CTRL_1_OFFSET, PORT_PRIO_QUEUE_ENABLE);
-}
-
/**
* sw_dis_diffserv - disable switch DiffServ priority
* @hw: The hardware instance.
@@ -2694,23 +2544,6 @@ static void sw_cfg_port_base_vlan(struct ksz_hw *hw, int port, u8 member)
}
/**
- * sw_get_addr - get the switch MAC address.
- * @hw: The hardware instance.
- * @mac_addr: Buffer to store the MAC address.
- *
- * This function retrieves the MAC address of the switch.
- */
-static inline void sw_get_addr(struct ksz_hw *hw, u8 *mac_addr)
-{
- int i;
-
- for (i = 0; i < 6; i += 2) {
- mac_addr[i] = readb(hw->io + KS8842_MAC_ADDR_0_OFFSET + i);
- mac_addr[1 + i] = readb(hw->io + KS8842_MAC_ADDR_1_OFFSET + i);
- }
-}
-
-/**
* sw_set_addr - configure switch MAC address
* @hw: The hardware instance.
* @mac_addr: The MAC address.
@@ -2917,56 +2750,6 @@ static inline void hw_w_phy_ctrl(struct ksz_hw *hw, int phy, u16 data)
writew(data, hw->io + phy + KS884X_PHY_CTRL_OFFSET);
}
-static inline void hw_r_phy_link_stat(struct ksz_hw *hw, int phy, u16 *data)
-{
- *data = readw(hw->io + phy + KS884X_PHY_STATUS_OFFSET);
-}
-
-static inline void hw_r_phy_auto_neg(struct ksz_hw *hw, int phy, u16 *data)
-{
- *data = readw(hw->io + phy + KS884X_PHY_AUTO_NEG_OFFSET);
-}
-
-static inline void hw_w_phy_auto_neg(struct ksz_hw *hw, int phy, u16 data)
-{
- writew(data, hw->io + phy + KS884X_PHY_AUTO_NEG_OFFSET);
-}
-
-static inline void hw_r_phy_rem_cap(struct ksz_hw *hw, int phy, u16 *data)
-{
- *data = readw(hw->io + phy + KS884X_PHY_REMOTE_CAP_OFFSET);
-}
-
-static inline void hw_r_phy_crossover(struct ksz_hw *hw, int phy, u16 *data)
-{
- *data = readw(hw->io + phy + KS884X_PHY_CTRL_OFFSET);
-}
-
-static inline void hw_w_phy_crossover(struct ksz_hw *hw, int phy, u16 data)
-{
- writew(data, hw->io + phy + KS884X_PHY_CTRL_OFFSET);
-}
-
-static inline void hw_r_phy_polarity(struct ksz_hw *hw, int phy, u16 *data)
-{
- *data = readw(hw->io + phy + KS884X_PHY_PHY_CTRL_OFFSET);
-}
-
-static inline void hw_w_phy_polarity(struct ksz_hw *hw, int phy, u16 data)
-{
- writew(data, hw->io + phy + KS884X_PHY_PHY_CTRL_OFFSET);
-}
-
-static inline void hw_r_phy_link_md(struct ksz_hw *hw, int phy, u16 *data)
-{
- *data = readw(hw->io + phy + KS884X_PHY_LINK_MD_OFFSET);
-}
-
-static inline void hw_w_phy_link_md(struct ksz_hw *hw, int phy, u16 data)
-{
- writew(data, hw->io + phy + KS884X_PHY_LINK_MD_OFFSET);
-}
-
/**
* hw_r_phy - read data from PHY register
* @hw: The hardware instance.
--
1.7.10.4
^ permalink raw reply related
* [PATCH] net: fddi: skfp: smt.c: Remove unused function
From: Rickard Strandqvist @ 2015-01-01 17:01 UTC (permalink / raw)
To: netdev, linux-kernel; +Cc: Rickard Strandqvist
Remove the function smt_ifconfig() that is not used anywhere.
This was partially found by using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
---
drivers/net/fddi/skfp/smt.c | 12 ------------
1 file changed, 12 deletions(-)
diff --git a/drivers/net/fddi/skfp/smt.c b/drivers/net/fddi/skfp/smt.c
index 9edada8..cd78b7c 100644
--- a/drivers/net/fddi/skfp/smt.c
+++ b/drivers/net/fddi/skfp/smt.c
@@ -1736,18 +1736,6 @@ char *addr_to_string(struct fddi_addr *addr)
}
#endif
-#ifdef AM29K
-int smt_ifconfig(int argc, char *argv[])
-{
- if (argc >= 2 && !strcmp(argv[0],"opt_bypass") &&
- !strcmp(argv[1],"yes")) {
- smc->mib.fddiSMTBypassPresent = 1 ;
- return 0;
- }
- return amdfddi_config(0, argc, argv);
-}
-#endif
-
/*
* return static mac index
*/
--
1.7.10.4
^ permalink raw reply related
* Re: [PATCH iproute2 v2] bridge/link: add learning_sync policy flag
From: Stephen Hemminger @ 2015-01-01 18:03 UTC (permalink / raw)
To: sfeldma; +Cc: netdev, jiri, roopa
In-Reply-To: <1419884407-21998-1-git-send-email-sfeldma@gmail.com>
On Mon, 29 Dec 2014 12:20:07 -0800
sfeldma@gmail.com wrote:
> From: Scott Feldman <sfeldma@gmail.com>
>
> v2:
>
> Resending now that the dust has cleared in 3.18 on "self" vs. hwmode debate for
> brport settings. learning_sync is now set/cleared using "self" qualifier on
> brport.
>
> v1:
>
> Add 'learned_sync' flag to turn on/off syncing of learned MAC addresses from
> offload device to bridge's FDB. Flag is be set/cleared on offload device port
> using "self" qualifier:
>
> $ sudo bridge link set dev swp1 learning_sync on self
>
> $ bridge -d link show dev swp1
> 2: swp1 state UNKNOWN : <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 master br0 state forwarding priority 32 cost 2
> hairpin off guard off root_block off fastleave off learning off flood off
> 2: swp1 state UNKNOWN : <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 master br0
> learning on learning_sync on
>
> Adds new IFLA_BRPORT_LEARNED_SYNCED attribute for IFLA_PROTINFO on the SELF
> brport.
>
> Signed-off-by: Scott Feldman <sfeldma@gmail.com>
Applied
^ permalink raw reply
* Re: [PATCH net-next v1 2/5] ixgbevf: Add a RETA query code
From: Alexander Duyck @ 2015-01-01 18:09 UTC (permalink / raw)
To: Vlad Zolotarov, netdev
Cc: gleb, avi, jeffrey.t.kirsher, Don Skidmore, tantilov, Emil S
In-Reply-To: <54A44188.204@cloudius-systems.com>
On 12/31/2014 10:33 AM, Vlad Zolotarov wrote:
>
> On 12/31/14 20:00, Alexander Duyck wrote:
>> I suspect this code is badly broken as it doesn't take several things
>> into account.
>>
>> First the PF redirection table can have values outside of the range
>> supported by the VF. This is allowed as the VF can set how many bits of
>> the redirection table it actually wants to use. This is controlled via
>> the PSRTYPE register. So for example the PF can be running with 4
>> queues, and the VF can run either in single queue or as just a pair of
>> queues.
>>
>> Second you could compress this data much more tightly by taking
>> advantage of the bit widths allowed. So for everything x540 and older
>> they only use a 4 bit value per entry. That means you could
>> theoretically stuff 8 entries per u32 instead of just 4.
>
> Compression is nice but I think ethtool expects it in a certain
> format: one entry per byte. And since this patch is targeting the
> ethtool the output format should be as ethtool expects it to be and
> this is what this patch does. However I agree that masking the
> appropriate bits according to PSRTYPE is required. Good catch!
The idea of compression comes into play when you consider there is
significant latency trying to get messages across the mailbox. By
reducing the number of messages needed to get the redirection table you
should be able to significantly reduce the amount of time needed to
fetch it. The job of compressing/expanding the values is actually
pretty straight forward when you consider all that should be needed is a
simple loop to perform some shift, and, and or operations.
- Alex
^ permalink raw reply
* Re: [PATCH] Drivers: isdn: gigaset: checkpatch cleanup
From: Tilman Schmidt @ 2015-01-01 18:46 UTC (permalink / raw)
To: Bas Peters, hjlipp; +Cc: isdn, gigaset307x-common, netdev, linux-kernel
In-Reply-To: <1420047298-7798-1-git-send-email-baspeters93@gmail.com>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hello Bas,
I have several objections to your patch.
Am 31.12.2014 um 18:34 schrieb Bas Peters:
> I have not been able to test the code as I do not have access to
> the hardware but since no new features were really added I don't
> think that should pose a problem.
It's always problematic to change code you cannot test.
At the very least, if you do coding style cleanups you should test
whether the result still compiles and generates the same code as before.
> --- a/drivers/isdn/gigaset/bas-gigaset.c +++
> b/drivers/isdn/gigaset/bas-gigaset.c @@ -261,11 +261,12 @@ static
> inline void dump_urb(enum debuglevel level, const char *tag, {
> #ifdef CONFIG_GIGASET_DEBUG int i; + gig_dbg(level, "%s
> urb(0x%08lx)->{", tag, (unsigned long) urb); if (urb) {
> gig_dbg(level, - " dev=0x%08lx, pipe=%s:EP%d/DV%d:%s, " -
> "hcpriv=0x%08lx, transfer_flags=0x%x,", + " dev=0x%08lx,
> pipe=%s:EP%d/DV%d:%s, + hcpriv=0x%08lx, transfer_flags=0x%x,",
This is syntactically wrong and won't compile. You cannot have an
unescaped newline inside a string literal.
> @@ -2312,13 +2312,13 @@ static int gigaset_probe(struct
> usb_interface *interface, /* Reject application specific
> interfaces */ if (hostif->desc.bInterfaceClass != 255) { -
> dev_warn(&udev->dev, "%s: bInterfaceClass == %d\n", +
> dev_warn(&udev->dev, "%s: bInterfaceClass == %d\n",\ __func__,
> hostif->desc.bInterfaceClass); return -ENODEV; }
>
> dev_info(&udev->dev, - "%s: Device matched (Vendor: 0x%x,
> Product: 0x%x)\n", + "%s: Device matched (Vendor: 0x%x, Product:
> 0x%x)\n",\ __func__, le16_to_cpu(udev->descriptor.idVendor),
> le16_to_cpu(udev->descriptor.idProduct));
This looks strange, and not like correct coding style. Why would you
want to escape the end of line after a function argument?
> --- a/drivers/isdn/gigaset/common.c +++
> b/drivers/isdn/gigaset/common.c @@ -53,7 +53,7 @@ void
> gigaset_dbg_buffer(enum debuglevel level, const unsigned char
> *msg, { unsigned char outbuf[80]; unsigned char c; - size_t space =
> sizeof outbuf - 1; + size_t space = sizeof(outbuf - 1);
This is wrong. The sizeof operator must be applied to the array
variable outbuf, not to the expression (outbuf - 1).
> --- a/drivers/isdn/gigaset/ev-layer.c +++
> b/drivers/isdn/gigaset/ev-layer.c
> @@ -1355,8 +1351,20 @@ static void do_action(int action, struct
> cardstate *cs, }
>
> for (i = 0; i < 4; ++i) { - val = simple_strtoul(s, (char **) &e,
> 10); - if (val > INT_MAX || e == s) + unsigned long *e; + +
> val = kstrtoul(s, 10, e); + if (val == -EINVAL) { +
> dev_err(cs->dev, "Parsing error on converting string to\ +
> unsigned long\n"); + break; + } + if (val == -ERANGE) { +
> dev_err(cs->dev, "Overflow error converting string to\ +
> unsigned long\n"); + break; + } + if (val > INT_MAX || *e ==
> s) break; if (i == 3) { if (*e)
This cannot work. The pointer variable e gets dereferenced without
ever being initialized. The type mismatches when declaring e as
pointing to an unsigned long but comparing *e to s in one place and to
a character literal in another point make me wonder which semantics
you had in mind for e in the first place.
Also your error messages are not helpful for someone reading the log
and trying to find out what went wrong, and not very readable because
of the big stretch of whitespace you insert between the words "to" and
"unsigned". In fact I'm not even convinced it's a good idea to emit a
log message at all here.
> --- a/drivers/isdn/gigaset/gigaset.h +++
> b/drivers/isdn/gigaset/gigaset.h @@ -94,8 +94,7 @@ enum debuglevel
> { #define gig_dbg(level, format, arg...) \ do { \ if
> (unlikely(((enum debuglevel)gigaset_debuglevel) & (level))) \ -
> printk(KERN_DEBUG KBUILD_MODNAME ": " format "\n", \ - ##
> arg); \ + dev_dbg(cs->dev, KBUILD_MODNAME ": " format "\n")\
> } while (0) #define DEBUG_DEFAULT (DEBUG_TRANSCMD | DEBUG_CMD |
> DEBUG_USBREQ)
>
This will not work when
- - there is no cs variable in the context where the macro is used or
- - cs->dev doesn't contain a valid device pointer or
- - the format string references additional arguments,
all of which actually occur in the driver.
> --- a/drivers/isdn/gigaset/i4l.c +++ b/drivers/isdn/gigaset/i4l.c
> @@ -624,14 +624,14 @@ int gigaset_isdn_regdev(struct cardstate *cs,
> const char *isdnid) { isdn_if *iif;
>
> - iif = kmalloc(sizeof *iif, GFP_KERNEL); + iif =
> kmalloc(sizeof(*iif, GFP_KERNEL)); if (!iif) { pr_err("out of
> memory\n"); return -ENOMEM;
You're calling kmalloc with too few arguments here.
> --- a/drivers/isdn/gigaset/proc.c +++
> b/drivers/isdn/gigaset/proc.c @@ -27,13 +27,18 @@ static ssize_t
> set_cidmode(struct device *dev, struct device_attribute *attr,
> const char *buf, size_t count) { struct cardstate *cs =
> dev_get_drvdata(dev); - long int value; - char *end; + long int
> *value; + int result;
>
> - value = simple_strtol(buf, &end, 0); - while (*end) - if
> (!isspace(*end++)) - return -EINVAL; + result = kstrtol(buf, 0,
> &value); + if (result == -ERANGE) + /* Overflow error */ +
> dev_err(cs->dev, "Overflow error on conversion from string to\ +
> long\n"); + if (result == -EINVAL) + /* Parsing error */ +
> dev_err(cs->dev, "Parsing error on conversion from string to\ +
> long\n"); if (value < 0 || value > 1) return -EINVAL;
This changes semantics. Your code will not accept the same input as
the original code, and it will emit messages of its own instead of
just returning an error code to the caller as it should.
> --- a/drivers/isdn/gigaset/usb-gigaset.c +++
> b/drivers/isdn/gigaset/usb-gigaset.c
> default: rate = 9600; - dev_err(cs->dev, "unsupported baudrate
> request 0x%x," - " using default of B9600\n", cflag); +
> dev_err(cs->dev, "unsupported baudrate request 0x%x,\ + using
> default of B9600\n", cflag);
This makes the message much less readable by inserting a long stretch
of whitespace after the comma.
In sum: NACK.
Regards,
Tilman
- --
Tilman Schmidt E-Mail: tilman@imap.cc
Bonn, Germany
Diese Nachricht besteht zu 100% aus wiederverwerteten Bits.
Ungeöffnet mindestens haltbar bis: (siehe Rückseite)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAEBAgAGBQJUpZYFAAoJEFPuqx0v+F+qEcsH/1yyu8A8tHPhiW60DzQWhCj7
kxKw7gUS24ATLEEl5jUmrua2xPM0Exg7FknBSYRmNmOEj8j3sl7mQ0dDzDcJgOgI
BaDXV5YqnnqppmYkdT7OMykAuhdt2rk1w4khc2EjyyKrAdGJyB+j3ROgRDtG0wsY
zI/Uz7yKe540cwVWc6VCNQvS7cVEasQZnJzzTGBcPW35RjTYpvWEieJ/yY3tphIe
TQRl+SrKgiwGuzi0p886Vk8Mu4cfHHO5/EXyzpVdMVg6wxwxNs+YeW5xRf/mjzQe
YyygRKUA4VtZTno/rabdA2QvLkdDMHoiUNYa0InmBpAlLVol8OLh5mTit9yozwY=
=uU+S
-----END PGP SIGNATURE-----
^ permalink raw reply
* Re: [PATCH] TCP: Add support for TCP Stealth
From: Stephen Hemminger @ 2015-01-01 19:06 UTC (permalink / raw)
To: Julian Kirsch; +Cc: netdev, Christian Grothoff, Jacob Appelbaum
In-Reply-To: <54A470B3.3010501@sec.in.tum.de>
On Wed, 31 Dec 2014 22:54:59 +0100
Julian Kirsch <kirschju@sec.in.tum.de> wrote:
> +#ifdef CONFIG_TCP_STEALTH
> +/* Stealth TCP socket configuration */
> + struct {
> + #define TCP_STEALTH_MODE_AUTH BIT(0)
> + #define TCP_STEALTH_MODE_INTEGRITY BIT(1)
> + #define TCP_STEALTH_MODE_INTEGRITY_LEN BIT(2)
> + int mode;
> + u8 secret[MD5_MESSAGE_BYTES];
> + int integrity_len;
> + u16 integrity_hash;
> + struct skb_mstamp mstamp;
> + bool saw_tsval;
> + } stealth;
> +#endif
If you want a bitfield, why not use a bitfield for mode?
If you have to use masks, better to use u8 for mode.
Integrity length should be unsigned since obviously negative
values are not possible.
Rearrange structure to save space. Lots of holes here.
^ permalink raw reply
* Re: [PATCH] TCP: Add support for TCP Stealth
From: Stephen Hemminger @ 2015-01-01 19:10 UTC (permalink / raw)
To: Julian Kirsch; +Cc: netdev, Christian Grothoff, Jacob Appelbaum
In-Reply-To: <54A470B3.3010501@sec.in.tum.de>
On Wed, 31 Dec 2014 22:54:59 +0100
Julian Kirsch <kirschju@sec.in.tum.de> wrote:
> + memcpy(iv, (const __u8 *)daddr,
> + (daddr_size > sizeof(iv)) ? sizeof(iv) : daddr_size);
> +
> +#ifdef CONFIG_TCP_MD5SIG
> + md5 = tp->af_specific->md5_lookup(sk, sk);
> +#else
> + md5 = NULL;
> +#endif
> + if (likely(sysctl_tcp_timestamps && !md5) || tp->stealth.saw_tsval)
> + tsval = tp->stealth.mstamp.stamp_jiffies;
> +
> + ((__be16 *)iv)[2] ^= cpu_to_be16(tp->stealth.integrity_hash);
Cast unnecessary on memcpy arg since it takes void *
Would be clearer to use a real structure or union not assignment to cast to setup iv.
^ permalink raw reply
* [PATCH] isdn: hisax: hfc4s8s_l1: Remove some unused functions
From: Rickard Strandqvist @ 2015-01-01 19:17 UTC (permalink / raw)
To: Karsten Keil, David S. Miller
Cc: Rickard Strandqvist, Dan Carpenter, netdev, linux-kernel
Removes some functions that are not used anywhere:
Read_hfc32() Write_hfc32() Write_hfc16()
This was partially found by using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
---
drivers/isdn/hisax/hfc4s8s_l1.c | 21 ---------------------
1 file changed, 21 deletions(-)
diff --git a/drivers/isdn/hisax/hfc4s8s_l1.c b/drivers/isdn/hisax/hfc4s8s_l1.c
index fc9f9d0..0e5d673 100644
--- a/drivers/isdn/hisax/hfc4s8s_l1.c
+++ b/drivers/isdn/hisax/hfc4s8s_l1.c
@@ -225,20 +225,6 @@ fWrite_hfc8(hfc4s8s_hw *a, u_char c)
}
static inline void
-Write_hfc16(hfc4s8s_hw *a, u_char b, u_short c)
-{
- SetRegAddr(a, b);
- outw(c, a->iobase);
-}
-
-static inline void
-Write_hfc32(hfc4s8s_hw *a, u_char b, u_long c)
-{
- SetRegAddr(a, b);
- outl(c, a->iobase);
-}
-
-static inline void
fWrite_hfc32(hfc4s8s_hw *a, u_long c)
{
outl(c, a->iobase);
@@ -266,13 +252,6 @@ Read_hfc16(hfc4s8s_hw *a, u_char b)
}
static inline u_long
-Read_hfc32(hfc4s8s_hw *a, u_char b)
-{
- SetRegAddr(a, b);
- return (inl((volatile u_int) a->iobase));
-}
-
-static inline u_long
fRead_hfc32(hfc4s8s_hw *a)
{
return (inl((volatile u_int) a->iobase));
--
1.7.10.4
^ permalink raw reply related
* Re: [PATCH net-next v2 2/2] bridge: modify bridge af spec parser to accomodate vlan list and ranges
From: Scott Feldman @ 2015-01-01 19:20 UTC (permalink / raw)
To: Roopa Prabhu; +Cc: Netdev, hemminger, vyasevic@redhat.com, Wilson Kok
In-Reply-To: <1420044533-16963-3-git-send-email-roopa@cumulusnetworks.com>
On Wed, Dec 31, 2014 at 8:48 AM, <roopa@cumulusnetworks.com> wrote:
> From: Roopa Prabhu <roopa@cumulusnetworks.com>
>
> This patch modifies br_afspec to parse incoming IFLA_BRIDGE_VLAN_INFO_LIST
>
> Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
> ---
> net/bridge/br_netlink.c | 115 ++++++++++++++++++++++++++++++++++-------------
> 1 file changed, 85 insertions(+), 30 deletions(-)
>
> diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
> index 492ef6a..bcba9d2 100644
> --- a/net/bridge/br_netlink.c
> +++ b/net/bridge/br_netlink.c
> @@ -226,53 +226,108 @@ static const struct nla_policy ifla_br_policy[IFLA_MAX+1] = {
> [IFLA_BRIDGE_VLAN_INFO_LIST] = { .type = NLA_NESTED, },
> };
>
> +static int br_vlan_info(struct net_bridge *br, struct net_bridge_port *p,
> + int cmd, struct bridge_vlan_info *vinfo)
> +{
> + int err = 0;
> +
> + switch (cmd) {
> + case RTM_SETLINK:
> + if (p) {
> + err = nbp_vlan_add(p, vinfo->vid, vinfo->flags);
> + if (err)
> + break;
> +
> + if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
> + err = br_vlan_add(p->br, vinfo->vid,
> + vinfo->flags);
> + } else {
> + err = br_vlan_add(br, vinfo->vid, vinfo->flags);
> + }
> + break;
> +
> + case RTM_DELLINK:
> + if (p) {
> + nbp_vlan_delete(p, vinfo->vid);
> + if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
> + br_vlan_delete(p->br, vinfo->vid);
> + } else {
> + br_vlan_delete(br, vinfo->vid);
> + }
> + break;
> + }
> +
> + return err;
> +}
> +
> static int br_afspec(struct net_bridge *br,
> struct net_bridge_port *p,
> struct nlattr *af_spec,
> int cmd)
> {
> struct nlattr *tb[IFLA_BRIDGE_MAX+1];
> + struct nlattr *attr;
> int err = 0;
> + int rem;
>
> err = nla_parse_nested(tb, IFLA_BRIDGE_MAX, af_spec, ifla_br_policy);
> if (err)
> return err;
>
> if (tb[IFLA_BRIDGE_VLAN_INFO]) {
> - struct bridge_vlan_info *vinfo;
> -
> - vinfo = nla_data(tb[IFLA_BRIDGE_VLAN_INFO]);
> -
> - if (!vinfo->vid || vinfo->vid >= VLAN_VID_MASK)
> - return -EINVAL;
> -
> - switch (cmd) {
> - case RTM_SETLINK:
> - if (p) {
> - err = nbp_vlan_add(p, vinfo->vid, vinfo->flags);
> - if (err)
> - break;
> -
> - if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
> - err = br_vlan_add(p->br, vinfo->vid,
> - vinfo->flags);
> - } else
> - err = br_vlan_add(br, vinfo->vid, vinfo->flags);
> -
> - break;
> -
> - case RTM_DELLINK:
> - if (p) {
> - nbp_vlan_delete(p, vinfo->vid);
> - if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
> - br_vlan_delete(p->br, vinfo->vid);
> - } else
> - br_vlan_delete(br, vinfo->vid);
> - break;
> + attr = tb[IFLA_BRIDGE_VLAN_INFO];
> + if (nla_len(attr) != sizeof(struct bridge_vlan_info))
> + goto err_inval;
Size check isn't necessary here as it was already done with nla_parse_nested().
> +
> + err = br_vlan_info(br, p, cmd,
> + (struct bridge_vlan_info *)nla_data(attr));
> +
> + } else if (tb[IFLA_BRIDGE_VLAN_INFO_LIST]) {
Could you have both IFLA_BRIDGE_VLAN_INFO and
IFLA_BRIDGE_VLAN_INFO_LIST? If so, drop the else.
> + struct bridge_vlan_info *vinfo_start = NULL;
> + struct bridge_vlan_info *vinfo = NULL;
Initializer not needed on this ^^^ one.
> +
> + nla_for_each_nested(attr, tb[IFLA_BRIDGE_VLAN_INFO_LIST], rem) {
> + if (nla_len(attr) != sizeof(struct bridge_vlan_info) ||
> + nla_type(attr) != IFLA_BRIDGE_VLAN_INFO)
> + goto err_inval;
goto isn't necessary...just return -EINVAL...more like this below
> + vinfo = nla_data(attr);
> + if (vinfo->flags & BRIDGE_VLAN_INFO_RANGE_START) {
> + if (vinfo_start)
> + goto err_inval;
> + vinfo_start = vinfo;
> + continue;
> + }
> +
> + if (vinfo_start) {
> + int v;
> +
> + if (!(vinfo->flags & BRIDGE_VLAN_INFO_RANGE_END))
> + goto err_inval;
> +
> + if (vinfo->vid < vinfo_start->vid)
> + goto err_inval;
> +
> + for (v = vinfo_start->vid; v <= vinfo->vid;
> + v++) {
Did the above exceed 80 chars? If not, one liner.
> + vinfo_start->vid = v;
> + err = br_vlan_info(br, p, cmd,
> + vinfo_start);
> + if (err)
> + break;
> + }
> + vinfo_start = NULL;
> + } else {
> + err = br_vlan_info(br, p, cmd, vinfo);
> + }
> + if (err)
> + break;
> }
> }
>
> return err;
> +
> +err_inval:
> + return -EINVAL;
> }
>
> static const struct nla_policy br_port_policy[IFLA_BRPORT_MAX + 1] = {
> --
> 1.7.10.4
>
^ permalink raw reply
* Re: [PATCH net-next v2 2/2] bridge: modify bridge af spec parser to accomodate vlan list and ranges
From: Scott Feldman @ 2015-01-01 19:34 UTC (permalink / raw)
To: Arad, Ronen
Cc: roopa@cumulusnetworks.com, netdev@vger.kernel.org,
hemminger@vyatta.com, vyasevic@redhat.com,
wkok@cumulusnetworks.com
In-Reply-To: <E4CD12F19ABA0C4D8729E087A761DC3505DD2DE4@ORSMSX101.amr.corp.intel.com>
On Thu, Jan 1, 2015 at 12:54 AM, Arad, Ronen <ronen.arad@intel.com> wrote:
>
>
>>-----Original Message-----
>>From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org] On
>>Behalf Of roopa@cumulusnetworks.com
>>Sent: Wednesday, December 31, 2014 6:49 PM
>>To: netdev@vger.kernel.org; hemminger@vyatta.com; vyasevic@redhat.com
>>Cc: sfeldma@gmail.com; wkok@cumulusnetworks.com; Roopa Prabhu
>>Subject: [PATCH net-next v2 2/2] bridge: modify bridge af spec parser to
>>accomodate vlan list and ranges
>>
>>From: Roopa Prabhu <roopa@cumulusnetworks.com>
>>
>>This patch modifies br_afspec to parse incoming IFLA_BRIDGE_VLAN_INFO_LIST
>>
>>Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
>>---
>> net/bridge/br_netlink.c | 115 ++++++++++++++++++++++++++++++++++------------
>>-
>> 1 file changed, 85 insertions(+), 30 deletions(-)
>>
>>diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
>>index 492ef6a..bcba9d2 100644
>>--- a/net/bridge/br_netlink.c
>>+++ b/net/bridge/br_netlink.c
>>@@ -226,53 +226,108 @@ static const struct nla_policy
>>ifla_br_policy[IFLA_MAX+1] = {
>> [IFLA_BRIDGE_VLAN_INFO_LIST] = { .type = NLA_NESTED, },
>> };
>>
>>+static int br_vlan_info(struct net_bridge *br, struct net_bridge_port *p,
>>+ int cmd, struct bridge_vlan_info *vinfo)
>>+{
>>+ int err = 0;
>>+
>>+ switch (cmd) {
>>+ case RTM_SETLINK:
>>+ if (p) {
>>+ err = nbp_vlan_add(p, vinfo->vid, vinfo->flags);
>>+ if (err)
>>+ break;
>>+
>>+ if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
>>+ err = br_vlan_add(p->br, vinfo->vid,
>>+ vinfo->flags);
>>+ } else {
>>+ err = br_vlan_add(br, vinfo->vid, vinfo->flags);
>>+ }
>>+ break;
>>+
>>+ case RTM_DELLINK:
>>+ if (p) {
>>+ nbp_vlan_delete(p, vinfo->vid);
>>+ if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
>>+ br_vlan_delete(p->br, vinfo->vid);
>>+ } else {
>>+ br_vlan_delete(br, vinfo->vid);
>>+ }
>>+ break;
>>+ }
>>+
>>+ return err;
>>+}
>>+
>> static int br_afspec(struct net_bridge *br,
>> struct net_bridge_port *p,
>> struct nlattr *af_spec,
>> int cmd)
>> {
>> struct nlattr *tb[IFLA_BRIDGE_MAX+1];
>>+ struct nlattr *attr;
>> int err = 0;
>>+ int rem;
>>
>> err = nla_parse_nested(tb, IFLA_BRIDGE_MAX, af_spec, ifla_br_policy);
>> if (err)
>> return err;
>>
>> if (tb[IFLA_BRIDGE_VLAN_INFO]) {
>>- struct bridge_vlan_info *vinfo;
>>-
>>- vinfo = nla_data(tb[IFLA_BRIDGE_VLAN_INFO]);
>>-
>>- if (!vinfo->vid || vinfo->vid >= VLAN_VID_MASK)
>>- return -EINVAL;
>>-
>>- switch (cmd) {
>>- case RTM_SETLINK:
>>- if (p) {
>>- err = nbp_vlan_add(p, vinfo->vid, vinfo->flags);
>>- if (err)
>>- break;
>>-
>>- if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
>>- err = br_vlan_add(p->br, vinfo->vid,
>>- vinfo->flags);
>>- } else
>>- err = br_vlan_add(br, vinfo->vid, vinfo->flags);
>>-
>>- break;
>>-
>>- case RTM_DELLINK:
>>- if (p) {
>>- nbp_vlan_delete(p, vinfo->vid);
>>- if (vinfo->flags & BRIDGE_VLAN_INFO_MASTER)
>>- br_vlan_delete(p->br, vinfo->vid);
>>- } else
>>- br_vlan_delete(br, vinfo->vid);
>>- break;
>>+ attr = tb[IFLA_BRIDGE_VLAN_INFO];
>>+ if (nla_len(attr) != sizeof(struct bridge_vlan_info))
>>+ goto err_inval;
>>+
>>+ err = br_vlan_info(br, p, cmd,
>>+ (struct bridge_vlan_info *)nla_data(attr));
>>+
>>+ } else if (tb[IFLA_BRIDGE_VLAN_INFO_LIST]) {
>>+ struct bridge_vlan_info *vinfo_start = NULL;
>>+ struct bridge_vlan_info *vinfo = NULL;
>>+
>>+ nla_for_each_nested(attr, tb[IFLA_BRIDGE_VLAN_INFO_LIST], rem) {
>>+ if (nla_len(attr) != sizeof(struct bridge_vlan_info) ||
>>+ nla_type(attr) != IFLA_BRIDGE_VLAN_INFO)
>>+ goto err_inval;
>>+ vinfo = nla_data(attr);
>>+ if (vinfo->flags & BRIDGE_VLAN_INFO_RANGE_START) {
>>+ if (vinfo_start)
>>+ goto err_inval;
>>+ vinfo_start = vinfo;
>>+ continue;
>>+ }
>>+
>>+ if (vinfo_start) {
>>+ int v;
>>+
>>+ if (!(vinfo->flags & BRIDGE_VLAN_INFO_RANGE_END))
>>+ goto err_inval;
>>+
>>+ if (vinfo->vid < vinfo_start->vid)
>
> This check rejects inverted range. However it allows the RANGE_START and
> RANGE_END vinfos to have the same vid. Isn't it inconsistent with the rejection
> of a single vinfo with both RANGE_START and RANGE_END set?
Allowing both START and END to be set on single vinfo might simplify
the encoding of LIST, so maybe it should be allowed.
Roopa, I know you dropped the subsequent notification patches from the
set, but I suspect now with these new START/END markers, the
notification algorithm can be very close to the original loop, without
having to make copies of the vlan_bitmap and untagged_bitmap. Using
both START/END on a single vinfo will keep the loop simple for adding
single vids that are not in a range.
(hmmm...START/STOP or BEGIN/END? Seems START/END is mixing the two
concepts...BEGIN/END seems best)
^ permalink raw reply
* Re: [PATCH] drivers:isdn: Remove uneeded fix me comment in capi.c for the function,decode_ie
From: Tilman Schmidt @ 2015-01-01 21:18 UTC (permalink / raw)
To: Nicholas Krause, hjlipp; +Cc: isdn, gigaset307x-common, netdev, linux-kernel
In-Reply-To: <1420053714-24661-1-git-send-email-xerofoify@gmail.com>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Nicholas,
Am 31.12.2014 um 20:21 schrieb Nicholas Krause:
> Removes a no longer needed fix me comment for the
> function,decode_ie in the file, capi.c. This comment is no longer
> needed as the commit this was written in was 2009 when the new
> driver,three was introduced. Due to no breakage related to the
> newer version of this driver we can safely remove this comment.
I won't oppose removal of that comment. However your commit message is
not quite to the point.
The comment poses the question of whether the conversion to upper case
done by the toupper() calls in the two lines following is actually
necessary. The fact that there is no breakage with the current code
does not answer that question. It could only be answered by *removing*
the toupper() calls and *then* checking for breakage.
However the cost of two toupper() calls seems so small that it's
hardly worth the effort and risk and we should just leave them in.
Regards,
Tilman
> diff --git a/drivers/isdn/gigaset/capi.c
> b/drivers/isdn/gigaset/capi.c index ccec777..d55ba3f 100644 ---
> a/drivers/isdn/gigaset/capi.c +++ b/drivers/isdn/gigaset/capi.c @@
> -182,7 +182,6 @@ static void decode_ie(u8 *in, char *out) { int i =
> *in; while (i-- > 0) { - /* ToDo: conversion to upper case
> necessary? */ *out++ = toupper(hex_asc_hi(*++in)); *out++ =
> toupper(hex_asc_lo(*in)); }
>
- --
Tilman Schmidt E-Mail: tilman@imap.cc
Bonn, Germany
Diese Nachricht besteht zu 100% aus wiederverwerteten Bits.
Ungeöffnet mindestens haltbar bis: (siehe Rückseite)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAEBAgAGBQJUpbm7AAoJEFPuqx0v+F+qxFcH/1F6l/1ycTikNdwfsHesZW6Q
TXDO7YfguLDe+wyePEvHdBFW+AwEgGFWTz2EEqfpdu/dHwmWjZh8yTLkJEMfn4fz
jBDm2Kyw+uFzUSXPVCyMUNXgC7yATs6l2AEkX5FSr0BycUVyMIeemChB4Pf5DfQD
B1Lz86xKhM3HpdI/KIDXqqHK88mT/B+7+rUL6lyJ6GRU/168EetD5PzjXoQ67DUw
JaSpCTjf1nUQAL1Nw623QVT4le/vPXLhwpIIAhLmvwO3nn8XWx4WdiCGVm6bGgrV
coWwo8LRB/IMCwoZGh2RhmDy2nOYwj1k9NYLjrx+izhbQUSpwqoLUCZe6vhBTks=
=Ppmw
-----END PGP SIGNATURE-----
^ permalink raw reply
* Re: [PATCH] can: kvaser_usb: Don't free packets when tight on URBs
From: Stephen Hemminger @ 2015-01-01 21:59 UTC (permalink / raw)
To: Ahmed S. Darwish
Cc: Olivier Sobrie, Oliver Hartkopp, Wolfgang Grandegger,
Marc Kleine-Budde, David S. Miller, Paul Gortmaker, Linux-CAN,
netdev, LKML
In-Reply-To: <20141223154654.GB6460@vivalin-002>
On Tue, 23 Dec 2014 17:46:54 +0200
"Ahmed S. Darwish" <darwish.07@gmail.com> wrote:
> int ret = NETDEV_TX_OK;
> + bool kfree_skb_on_error = true;
>
> if (can_dropped_invalid_skb(netdev, skb))
> return NETDEV_TX_OK;
> @@ -1336,6 +1337,7 @@ static netdev_tx_t kvaser_usb_start_xmit(struct sk_buff *skb,
>
> if (!context) {
> netdev_warn(netdev, "cannot find free context\n");
> + kfree_skb_on_error = false;
> ret = NETDEV_TX_BUSY;
You already have a flag value (ret == NETDEV_TX_BUSY), why
not use that instead of introducing another variable?
^ permalink raw reply
* Re: [PATCH] TCP: Add support for TCP Stealth
From: Julian Kirsch @ 2015-01-01 23:31 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev, Christian Grothoff, Jacob Appelbaum
In-Reply-To: <20150101111030.1e2b3a18@urahara>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi Stephen,
thanks a lot for your input. I've reflected your suggestions in the
code: The mode member of the stealth struct is now a u8, integrity_len
is now unsigned (size_t), I've switched the integrity_len and
integrity_hash members in order to close the gaps in the struct (sorry
for wasting bits) and I removed the cast from the second parameter of
the memcopy you mentioned. While I totally agree with you that the
part of the iv-setup where we xor the hash, tsval and dport in is not
very readable, I'd argue that declaring and using a struct would make
the patch longer while the semantics of such a struct really are
needed only in context of a single function and therefore cannot be
reused. Do you think that accessing the elements with a macro (see
below) could be an alternative?
#define tcp_stealth_iv_integrity_hash(iv) (((__be16 *)&iv)[2])
Best,
Julian
On 2015-01-01 20:10, Stephen Hemminger wrote:
> On Wed, 31 Dec 2014 22:54:59 +0100 Julian Kirsch
> <kirschju@sec.in.tum.de> wrote:
>
>> + memcpy(iv, (const __u8 *)daddr, + (daddr_size >
>> sizeof(iv)) ? sizeof(iv) : daddr_size); + +#ifdef
>> CONFIG_TCP_MD5SIG + md5 = tp->af_specific->md5_lookup(sk, sk);
>> +#else + md5 = NULL; +#endif + if (likely(sysctl_tcp_timestamps
>> && !md5) || tp->stealth.saw_tsval) + tsval =
>> tp->stealth.mstamp.stamp_jiffies; + + ((__be16 *)iv)[2] ^=
>> cpu_to_be16(tp->stealth.integrity_hash);
>
> Cast unnecessary on memcpy arg since it takes void *
>
> Would be clearer to use a real structure or union not assignment to
> cast to setup iv. -- To unsubscribe from this list: send the line
> "unsubscribe netdev" in the body of a message to
> majordomo@vger.kernel.org More majordomo info at
> http://vger.kernel.org/majordomo-info.html
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAEBAgAGBQJUpdjhAAoJENwkOWttRRA4ZFYIALs7gskvZqlCzFCuNTsJ4js9
7x5OTsuyY5caOIEcveVqYnW2dcOO2Jtwe1QUIOsRo7X6YOEA/8IT6+sJ0fxViLTD
TJAzU670Kcecn7+0cHPAj31yW+t9SHb5BBzMLCJlhSAboMs0YKmkwetqffg013uP
x81OI6kJy6pUCAeBeyyy20QafrIhs5vjEILGf9qSzeoIXRBdpnuH99FzoxEjOkUA
ka4QtrAUh3Uk0s6H8ezcpqvY2bKcz7te8+af5XF+Kz/+DLatoN2x58psxw3irBCw
x18FRVsbgmXY/m3leKilK4ieCyO1LzafRNa674fSW6QtHHMDZwSoYi0kDjSNMRk=
=ke3l
-----END PGP SIGNATURE-----
^ permalink raw reply
* Re: [PATCH v3 0/6] support GMAC driver for RK3288
From: David Miller @ 2015-01-02 2:01 UTC (permalink / raw)
To: heiko
Cc: roger.chen, peppe.cavallaro, netdev, linux-kernel, linux-rockchip,
kever.yang, eddie.cai
In-Reply-To: <1493383.y3pPThNXY3@phil>
From: Heiko Stübner <heiko@sntech.de>
Date: Thu, 01 Jan 2015 03:06:31 +0100
> Hi David,
>
> Am Mittwoch, 31. Dezember 2014, 19:15:38 schrieb David Miller:
>> From: Roger Chen <roger.chen@rock-chips.com>
>> Date: Mon, 29 Dec 2014 17:42:32 +0800
>>
>> > Roger Chen (6):
>> > patch1: add driver for Rockchip RK3288 SoCs integrated GMAC
>> > patch2: define clock ID used for GMAC
>> > patch3: modify CRU config for Rockchip RK3288 SoCs integrated GMAC
>> > patch4: dts: rockchip: add gmac info for rk3288
>> > patch5: dts: rockchip: enable gmac on RK3288 evb board
>> > patch6: add document for Rockchip RK3288 GMAC
>> >
>> > Tested on rk3288 evb board:
>> > Execute the following command to enable ethernet,
>> > set local IP and ping a remote host.
>> >
>> > busybox ifconfig eth0 up
>> > busybox ifconfig eth0 192.168.1.111
>> > ping 192.168.1.1
>>
>> Series applied to net-next, thanks.
>
> could we split this up a bit instead?
Too late, what's in my tree is in the permanent commit history
and cannot be deleted.
^ permalink raw reply
* Re: [net-next PATCH 00/17] fib_trie: Reduce time spent in fib_table_lookup by 35 to 75%
From: David Miller @ 2015-01-02 2:08 UTC (permalink / raw)
To: alexander.duyck; +Cc: alexander.h.duyck, netdev
In-Reply-To: <54A4B1D4.1030506@gmail.com>
From: Alexander Duyck <alexander.duyck@gmail.com>
Date: Wed, 31 Dec 2014 18:32:52 -0800
> On 12/31/2014 03:46 PM, David Miller wrote:
>> This knocks about 35 cpu cycles off of a lookup that ends up using the
>> default route on sparc64. From about ~438 cycles to ~403.
>
> Did that 438 value include both fib_table_lookup and check_leaf? Just
> curious as the overall gain seems smaller than what I have been seeing
> on the x86 system I was testing with, but then again it could just be a
> sparc64 thing.
This is just a default run of my kbench_mod.ko from the net_test_tools
repo. You can try it as well on x86-86 or similar.
> I've started work on a second round of patches. With any luck they
> should be ready by the time the next net-next opens. My hope is to cut
> the look-up time by another 30 to 50%, though it will take some time as
> I have to go though and drop the leaf_info structure, and look at
> splitting the tnode in half to break the key/pos/bits and child pointer
> dependency chain which will hopefully allow for a significant reduction
> in memory read stalls.
I'm very much looking forward to this.
> I am also planning to take a look at addressing the memory waste that
> occurs on nodes larger than 256 bytes due to the way kmalloc allocates
> memory as powers of 2. I'm thinking I might try encouraging the growth
> of smaller nodes, and discouraging anything over 256 by implementing a
> "truesize" type logic that can be used in the inflate/halve functions so
> that the memory usage is more accurately reflected.
Wouldn't this result in a deeper tree? The whole point is to keep the
tree as shallow as possible to minimize the memory refs on a lookup
right?
^ permalink raw reply
* [PATCH net-next 0/3] swdev: add IPv4 routing offload
From: sfeldma @ 2015-01-02 3:29 UTC (permalink / raw)
To: netdev, jiri, john.fastabend, tgraf, jhs, andy, roopa
From: Scott Feldman <sfeldma@gmail.com>
This patch set adds L3 routing offload support for IPv4 routes. The idea is to
mirror routes installed in the kernel's FIB down to a hardware switch device to
offload the data forwarding path for L3. Only the data forwarding path is
intercepted. Control and management of the kernel's FIB remains with the
kernel.
A couple of new ndo ops (ndo_switch_fib_ipv4_add/del) are added to the swdev
model to add/remove FIB entries to/from the offload device. The ops are called
from the core IPv4 FIB code directly. Just before the FIB entry is installed
in the kernel's FIB, the swdev device driver gets a chance at the FIB entry
(assuming the swdev driver implements the new ndo ops). This is a synchronous
call in the RTM_NEWROUTE path, and the swdev has the option to fail the
install, which means the FIB entry is not installed in swdev or the kernel, and
the user is notified of the failure. The swdev driver also has the option to
return -EOPNOTSUPP to pass on the FIB entry, so it'll only be installed in the
kernel FIB.
The FIB flush path is modified also to call into the swdev driver to flush the
FIB entries from hardware.
The rocker swdev driver is updated to support these new ndo ops. Right now
rocker only supports IPv4 singlepath routes, but follow-on patches will add
IPv6 and ECMP support. Also, only unicast IPv4 routes are supported, but
follow-on patches will add multicast route support.
Testing was done in my simulated network envionment using VMs and the rocker
device. I'm using Quagga OSPFv2 for the routing protocol for automatic control
plane processing. No modifications to Quagga or netlink/iproute2 is required;
it just works.
One important metric is the time spent installing/removing FIB entries from the
kernel and the device. With these patches applied, I measured the wall time
required to install and remove 10K IPv4 routes. I used ip route add cmd in
batch mode to install static routes. I used the ip route flush cmd to delete
the routes. This is 10000 routes installed to the kernel's FIB and to the
swdev device's L3 tables. And then removed from each. The performance is less
than a second for each operation. This is on my simulated rocker device running
on a VM, so a real embedded CPU would probably do much better.
My batch has 10K lines of:
simp@simp:~$ head east
route add 16.0.0.0/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.1/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.2/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.3/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.4/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.5/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.6/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.7/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.8/32 nexthop via 11.0.0.2 dev swp1
route add 16.0.0.9/32 nexthop via 11.0.0.2 dev swp1
[...]
Install/removing routes:
simp@simp:~$ wc -l east
10000 east
simp@simp:~$ ip route show root 16/8 | wc -l
0
simp@simp:~$ time sudo ip --batch east
real 0m0.715s
user 0m0.092s
sys 0m0.388s
simp@simp:~$ ip route show root 16/8 | wc -l
10000
[At this point, 10K routes are installed in kernel and the device]
simp@simp:~$ time sudo ip route flush root 16/8
real 0m0.458s
user 0m0.000s
sys 0m0.284s
simp@simp:~$ ip route show root 16/8 | wc -l
0
[All gone]
Scott Feldman (3):
net: add IPv4 routing FIB support for swdev
net: call swdev fib del for flushed routes
rocker: implement IPv4 fib offloading
drivers/net/ethernet/rocker/rocker.c | 441 +++++++++++++++++++++++++++++++++-
include/linux/netdevice.h | 22 ++
include/net/switchdev.h | 18 ++
net/ipv4/fib_trie.c | 31 ++-
net/switchdev/switchdev.c | 89 +++++++
5 files changed, 592 insertions(+), 9 deletions(-)
--
1.7.10.4
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox