Commit Graph

435 Commits

Author SHA1 Message Date
Koen Vandeputte
0a59e91973 ar71xx/ath79: ag71xx: dont fetch the same var again
tx_size was just declared above and set to BIT(tx->order)
Use the declaration instead, which could avoid a pointer deref

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
(cherry picked from commit 6852eab58236e3991f5d73f211018dfef01f2a64)
2019-08-27 10:31:06 +02:00
Koen Vandeputte
db1a935e5c ar71xx: ag71xx: use base address value directly from the source
This brings the code more inline with the ath79 flavour

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-27 10:31:06 +02:00
Koen Vandeputte
8d78aa8eaf ar71xx: ag71xx: make use of managed dev API simplifying code
Backport of a4eef43a12 ("ath79: ag71xx: replace alloc_etherdev with devm_alloc_etherdev")
combined with the initial changes from John Crispin.

Simplifies the code a lot by using the Managed dev API.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-27 10:31:06 +02:00
Koen Vandeputte
c9b28b5703 ar71xx: ag71xx: replace ag71xx_get_phy_if_mode_name with phy_modes()
Backport of f73b2d64ed ("ath79: ag71xx: replace ag71xx_get_phy_if_mode_name() with phy_modes()")

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-27 10:31:06 +02:00
Koen Vandeputte
d143a0a4f6 ar71xx: ag71xx: pass correct device pointer to dma functions
Backport of 4eaa3626a8 ("ath79: ag71xx: pass correct device pointer to dma functions")

While 4.14 does not contain the warnings,
it still makes sense to use the proper pointers here.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-27 10:31:06 +02:00
Koen Vandeputte
42b42d2c9b ar71xx/ath79: ag71xx: get ring_mask consistent
All other instances of this identical declaration fetch the
value directly from the ring_order.

Also do it here.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-27 10:31:06 +02:00
Koen Vandeputte
576791b9b4 ar71xx: ag71xx: add missing register writes
These are added in ath79, but were not backported here

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-27 10:31:06 +02:00
Koen Vandeputte
62eae4805b ar71xx: ag71xx: init rings with GFP_KERNEL
ar71xx got lost during final rebase ..

Fixes: c8a6ce71e4 ("ar71xx/ath79: ag71xx: init rings with GFP_KERNEL")
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-09 18:52:21 +02:00
Koen Vandeputte
dda342abba ar71xx/ath79: ag71xx: fix sleep in atomic
When enabling atomic-sleep-debugging options in the kernel,
following splat is seen when disabling the interface (which happens on boot):

[   10.892878] eth0: link down
[   10.896788] BUG: sleeping function called from invalid context at net/core/dev.c:5563
[   10.904730] in_atomic(): 1, irqs_disabled(): 1, pid: 425, name: ip
[   10.911004] 2 locks held by ip/425:
[   10.914539]  #0:  (rtnl_mutex){....}, at: [<80377474>] rtnetlink_rcv_msg+0x2d8/0x380
[   10.922441]  #1:  (&(&ag->lock)->rlock){....}, at: [<80330158>] ag71xx_hw_disable+0x24/0x94
[   10.930976] CPU: 0 PID: 425 Comm: ip Not tainted 4.14.136 #0
[   10.936716] Stack : 805e0000 80589228 80557404 876998ec 80610000 80610000 87cdcafc 805b5327
[   10.945233]         80551534 000001a9 8061386c 87699ccc 87cfb180 00000001 876998a0 84f70903
[   10.953751]         00000000 00000000 80b00000 8769979c 6a7407fa 00000000 00000007 00000000
[   10.962270]         000000b7 16d0954a 000000b6 00000000 80000000 87cb658c 87cb65b0 00000001
[   10.970787]         8046f97c 87699ccc 87cfb180 87ff2810 00000003 802ce724 0806e098 80610000
[   10.979306]         ...
[   10.981797] Call Trace:
[   10.984287] [<8006cb0c>] show_stack+0x58/0x100
[   10.988814] [<800aab34>] ___might_sleep+0x100/0x120
[   10.993774] [<8035c434>] napi_disable+0x30/0xd8
[   10.998377] [<80330198>] ag71xx_hw_disable+0x64/0x94
[   11.003418] [<8033069c>] ag71xx_stop+0x24/0x38
[   11.007959] [<80359e30>] __dev_close_many+0xcc/0x104
[   11.013009] [<80362eac>] __dev_change_flags+0xc8/0x1ac
[   11.018227] [<80362fb8>] dev_change_flags+0x28/0x70
[   11.023182] [<80376890>] do_setlink+0x31c/0x91c
[   11.027786] [<80379360>] rtnl_newlink+0x3ec/0x7f8
[   11.032563] [<80377498>] rtnetlink_rcv_msg+0x2fc/0x380
[   11.037799] [<8039a734>] netlink_rcv_skb+0xd4/0x178
[   11.042754] [<80399d10>] netlink_unicast+0x168/0x250
[   11.047796] [<8039a2d4>] netlink_sendmsg+0x3d8/0x434
[   11.052841] [<8033f0e4>] ___sys_sendmsg+0x1dc/0x290
[   11.057794] [<80340140>] __sys_sendmsg+0x54/0x84
[   11.062495] [<8007212c>] syscall_common+0x34/0x58

This is caused by calling napi_disable() while holding the spinlock.

Fix it by omitting the spinlock, which is not required here
Extensively tested on GL-MiFi, RB-912 and RB-922 hardware

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-08-09 18:32:38 +02:00
Petr Štetiar
971d358579 ar71xx: ag71xx: update ethtool support
Remove references to broken and mostly deprecated phy_ethtool_ioctl, use
new {s,g}et_link_ksettings and add nway_reset which was previously
handled in phy_ethtool_ioctl.

Cc: John Crispin <john@phrozen.org>
Ref: https://bugs.openwrt.org/index.php?do=details&task_id=1982
Signed-off-by: Petr Štetiar <ynezz@true.cz>
2019-06-05 10:12:30 +02:00
Petr Štetiar
eafe0611d9 ar71xx: ag71xx: remove unused SIOCETHTOOL ioctl handling
This ioctl is currently routed through generic interface code:

  dev_ioctl
    dev_ethtool
      __ethtool_get_link_ksettings
        phy_ethtool_ioctl

Cc: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
2019-06-05 10:12:30 +02:00
Petr Štetiar
31ea08a64b ar71xx: ag71xx: Fix broken networking on some devices (FS#2177)
It was reported, that latest ar71xx builds have broken networking on
TP-Link TL-WPA8630 and Nanostation M5 XW devices and that by reverting
the offending commit, everything is back to normal.

Fixes: d3506d1 ("ar71xx: ag71xx: fix compile error when enabling debug")
Signed-off-by: Petr Štetiar <ynezz@true.cz>
2019-04-08 18:37:04 +02:00
Koen Vandeputte
d3506d1bc1 ar71xx: ag71xx: fix compile error when enabling debug
Starting from kernel 4.5, phy_id needs to be fetched from a different location.

not doing so results in this compile error:

drivers/net/ethernet/atheros/ag71xx/ag71xx_phy.c: In function 'ag71xx_phy_connect_multi':
drivers/net/ethernet/atheros/ag71xx/ag71xx_phy.c:133:35: error: 'struct mdio_device' has no member named 'phy_id'
    ag->mii_bus->mdio_map[phy_addr]->phy_id);
                                   ^
./include/linux/printk.h:137:18: note: in definition of macro 'no_printk'
    printk(fmt, ##__VA_ARGS__); \
                  ^~~~~~~~~~~
drivers/net/ethernet/atheros/ag71xx/ag71xx.h:72:27: note: in expansion of macro 'pr_debug'
 #define DBG(fmt, args...) pr_debug(fmt, ## args)
                           ^~~~~~~~
drivers/net/ethernet/atheros/ag71xx/ag71xx_phy.c:130:3: note: in expansion of macro 'DBG'
   DBG("%s: PHY found at %s, uid=%08x\n",
   ^~~
scripts/Makefile.build:326: recipe for target 'drivers/net/ethernet/atheros/ag71xx/ag71xx_phy.o' failed

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2019-03-05 13:19:43 +01:00
Milan Krstic
39ede7bc67 ar71xx: ag71xx: preserve port mirror flags during swconfig apply
The swconfig load operation always triggers 'apply' function which in
this driver currently clears port mirroring flags effectively undoing
port mirroring configuration.

This fix preserves port mirroring flags during apply.

Signed-off-by: Milan Krstic <milan.krstic@gmail.com>
2019-01-22 09:05:59 +01:00
Petr Štetiar
22f9afa118 ar71xx: ag71xx: Replace duplicate debugging code with simple function call
ag71xx_dump_regs is used in code several times, and is providing same
output.

Signed-off-by: Petr Štetiar <ynezz@true.cz>
2018-12-17 08:03:45 +01:00
Koen Vandeputte
6100655c8c ar71xx: fix typo in rb91x nand source
Fixes: 3b1ea0996f ("ar71xx: fix build error due to bad include")

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2018-08-24 09:18:13 +02:00
Koen Vandeputte
3b1ea0996f ar71xx: fix build error due to bad include
While "rawnand.h" is available in kernel 4.14,
the default for this target is kernel 4.9 in which "nand.h" should be used.

Add an extra check to include the correct file depending on kernel version

Fixes these build errors:

drivers/mtd/nand/ar934x_nfc.c:16:10: fatal error: linux/mtd/rawnand.h: No such file or directory
 #include <linux/mtd/rawnand.h>
          ^~~~~~~~~~~~~~~~~~~~~
compilation terminated.

Fixes: 318e19ba67 ("ar71xx: add v4.14 support")

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2018-08-22 13:11:59 +02:00
Koen Vandeputte
743654f30d ar71xx: add missing include for checking kernel version
Fixes these build errors:

arch/mips/ath79/mach-rb2011.c:20:5: error: "LINUX_VERSION_CODE" is not defined, evaluates to 0 [-Werror=undef]
 #if LINUX_VERSION_CODE < KERNEL_VERSION(4,14,0)
     ^~~~~~~~~~~~~~~~~~
arch/mips/ath79/mach-rb2011.c:20:26: error: "KERNEL_VERSION" is not defined, evaluates to 0 [-Werror=undef]
 #if LINUX_VERSION_CODE < KERNEL_VERSION(4,14,0)
                          ^~~~~~~~~~~~~~
arch/mips/ath79/mach-rb2011.c:20:40: error: missing binary operator before token "("
 #if LINUX_VERSION_CODE < KERNEL_VERSION(4,14,0)
                                        ^

Fixes: 318e19ba67 ("ar71xx: add v4.14 support")

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
2018-08-22 13:11:52 +02:00
John Crispin
318e19ba67 ar71xx: add v4.14 support
adds v4.14 patches for testing but leaves v4.9 as default for now.

Signed-off-by: John Crispin <john@phrozen.org>
2018-08-22 08:09:00 +02:00
David Bauer
f4f99ec973 ar71xx: fix QCA955X SGMII link loss
The QCA955X is affected by a hardware bug which causes link-loss of the
SGMII link between SoC and PHY. This happens on change of link-state or
speed.

It is not really known what causes this bug. It definitely occurs when
using a AR8033 Gigabit Ethernet PHY.

Qualcomm solves this Bug in a similar fashion. We need to apply the fix
on a per-device base via platform-data as performing the fixup work will
break connectivity in case the SGMII interface is connected to a Switch.

This bug was first proposed to be fixed by Sven Eckelmann in 2016.
 https://patchwork.ozlabs.org/patch/604782/

Based-on-patch-by: Sven Eckelmann <sven.eckelmann@open-mesh.com>
Signed-off-by: David Bauer <mail@david-bauer.net>
2018-08-08 15:48:57 +02:00
Jo-Philipp Wich
b9bd3285c7 Revert "ar71xx: ag71xx: Prevent kernel oops for board def"
This reverts commit 7a3e133751.

This change reportedly breaks connectivity on some ar71xx devices, so
revert it for now.

Ref: https://github.com/openwrt/openwrt/pull/1217#issuecomment-409708087
Reported-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
2018-08-01 22:34:22 +02:00
Jo-Philipp Wich
33dbe2c55d Revert "ar71xx: ag71xx: Add connect message: fixed phy"
This reverts commit 0b9f4e8808.

This change reportedly breaks connectivity on some ar71xx devices, so
revert it for now.

Ref: https://github.com/openwrt/openwrt/pull/1217#issuecomment-409708087
Reported-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
2018-08-01 22:31:37 +02:00
Jo-Philipp Wich
9ebf596386 Revert "ar71xx: ag71xx_phy: Fix compilation for debug messages"
This reverts commit 2655fbe8ef.

The patch introduces syntax errors, revert it for now.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
2018-08-01 22:31:14 +02:00
Daniel F. Dickinson
7a3e133751 ar71xx: ag71xx: Prevent kernel oops for board def
The driver is written in such a way that with a board defintion that
connects a fixed phy, mdio, and switch in a certain way, a kernel oops could
result because of lack of previously probed mdio bus.

This commit allows for easier debugging in this case by casting the
correct blame with serial console messages.

Signed-off-by: Daniel F. Dickinson <cshored@thecshore.com>
2018-07-30 10:43:36 +02:00
Daniel F. Dickinson
0b9f4e8808 ar71xx: ag71xx: Add connect message: fixed phy
It's a little noisier but makes it obvious when the ar7240 switch was
connected to the MDIO bus, and to which phy device (or the failure
to do so).

Signed-off-by: Daniel F. Dickinson <cshored@thecshore.com>
2018-07-30 10:43:35 +02:00
Daniel F. Dickinson
2655fbe8ef ar71xx: ag71xx_phy: Fix compilation for debug messages
NB: Error only appears with ag71xx debug messages and dynamic printk
enabled.  This is probably why no one has caught it before.

Previously phy probe debug messages used old (now wrong) functions
to get the phy name for printing.  There was also the chance of
a NULL pointer in the event no phy_device was found.

Signed-off-by: Daniel F. Dickinson <cshored@thecshore.com>
2018-07-30 10:43:35 +02:00
Felix Fietkau
83ddc3bfc2 ar71xx: fix mii bus irq assignment
Signed-off-by: Felix Fietkau <nbd@nbd.name>
2018-03-04 19:37:20 +01:00
Felix Fietkau
f62d2b9549 ar71xx: remove pdata->fifo_cfg*
The values are the same for all chipsets that use it

Signed-off-by: Felix Fietkau <nbd@nbd.name>
2018-03-04 19:37:20 +01:00
Felix Fietkau
36fb0697e2 ar71xx: fix build with ag71xx debugfs support enabled
Signed-off-by: Felix Fietkau <nbd@nbd.name>
2018-02-28 13:57:58 +01:00
Rosen Penev
810659a22c ag71xx: Move timestamp struct member outside of struct.
With this change, the timestamp variable is only used in ag71xx_check_dma_stuck. Small tx speedup.

Based on a Qualcomm commit. ag->timestamp = jiffies was not replaced with netif_trans_update(dev) because of this quote:

It should be noted that after this series several instances
of netif_trans_update() are useless (if they occur in
.ndo_start_xmit and driver doesn't set LLTX flag -- stack already
did an update).

From: http://lists.openwall.net/netdev/2016/05/03/87

Signed-off-by: Rosen Penev <rosenp@gmail.com>
2018-02-28 12:46:02 +01:00
Felix Fietkau
798497c66e ar71xx: use global timestamp for hang check
Shrink the size of struct ag71xx_buf to 8 bytes, which improves cache
footprint

Signed-off-by: Felix Fietkau <nbd@nbd.name>
2018-02-05 10:16:25 +01:00
Felix Fietkau
31e6016d2e ar71xx: only access device stats in tx handler if packets were processed
Signed-off-by: Felix Fietkau <nbd@nbd.name>
2018-02-05 10:16:25 +01:00
Felix Fietkau
890e7458b0 ar71xx: allocate rx/tx descriptor/buffers in one chunk
Reduces false sharing due to cache aliases

Signed-off-by: Felix Fietkau <nbd@nbd.name>
2018-02-05 10:16:25 +01:00
Rosen Penev
4e03a742e0 ag71xx: Reorder ag71xx struct members for better cache performance
Qualcomm claims this improves the D-cache footprint. Origina commit message below:

From: Ben Menchaca <ben.menchaca@qca.qualcomm.com>
Date: Fri, 7 Jun 2013 10:57:28 -0500
Subject: [ag71xx] cluster/align structs for cache perf

Cluster the frequently used, per-packet structures in ag71xx near
to each other, and cacheline-align them.  Some other re-ordering
occurred to move "warmer" structures near the per-packet structures.

Signed-off-by: Ben Menchaca <ben.menchaca@qca.qualcomm.com>
Signed-off-by: Rosen Penev <rosenp@gmail.com>
2018-02-05 10:16:25 +01:00
Gabor Juhos
8f44c8086b ar71xx: remove obsolete '< KERNEL_VERSION(4, 2, 0)' check
Because the oldest supported kernel version on the ar71xx target is 4.4,
the condition that the kernel version is smaller than 4.2.0 is always
false. Remove the obsolete check from ag71xx_main.c to clean up the code
a bit.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
2018-01-09 21:33:18 +01:00
Gabor Juhos
7a4930b27c ar71xx: remove obsolete '< KERNEL_VERSION(3, 15, 0)' checks
Because the oldest supported kernel version on the ar71xx target is 4.4,
the condition that the kernel version is smaller than 3.15.0 is always
false. Remove the obsolete checks from the target specific source files
to clean up the code a bit.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
2018-01-09 21:33:18 +01:00
Vittorio Gambaletta
d40a358136
ag71xx: Fix rx ring buffer stall on small packets flood on qca956x and qca953x.
Backported from Code Aurora QSDK

Signed-off-by: Vittorio Gambaletta <openwrt@vittgam.net>
2017-12-29 15:17:07 +01:00
Gabor Juhos
6456999731 ar71xx: spi-rb4xx: balance clk_prepare_enable() call
Commit c312cef223 ("ar71xx: spi-rb4xx fix.") replaced the clk_enable() call
with clk_prepare_enable() to meet the common clock framework requirements.
However it did not change the clk_disable() call in the error patch which
thus leads to imbalance.

Fix the code by using the correct counterpart of clk_prepare_enable() in both
places.

Fixes: c312cef223 ("ar71xx: spi-rb4xx fix.")
Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
2017-12-24 09:03:01 +01:00
Rosen Penev
1c2fdfbaf2 ag71xx: Reduce NAPI weight to 32.
Qualcomm claims this reduces cache misses. Original commit message below:

From: Ben Menchaca <ben.menchaca@qca.qualcomm.com>
Date: Tue, 11 Jun 2013 12:18:46 -0500
Subject: [ag71xx] reduce NAPI weight

In an attempt to increase our cache warmth, we are decreasing NAPI.
This increases the warmth of the reused SKBs.

Signed-off-by: Ben Menchaca <ben.menchaca@qca.qualcomm.com>
Signed-off-by: Rosen Penev <rosenp@gmail.com>
2017-12-08 19:54:37 +01:00
Rosen Penev
524d103e7e Revert "ag71xx: Switch from driver to kernel macro for NAPI_WEIGHT."
The motivation for this was misguided. It turns out tuning the NAPI weight could be useful for testing purposes. Therefore reverting.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
2017-12-08 19:54:34 +01:00
Rosen Penev
e555153807 ag71xx: Switch from driver to kernel macro for NAPI_WEIGHT.
NAPI_POLL_WEIGHT was introduced in the kernel for exactly this purpose 5 years ago.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
2017-11-06 16:39:41 +01:00
Felix Fietkau
7c09fa4a74 ar71xx: fix mikrotik routerboard nand driver issues with linux 4.9
The mtd device is now embedded inside the nand chip data structure

Signed-off-by: Felix Fietkau <nbd@nbd.name>
2017-10-24 13:24:04 +02:00
Felix Fietkau
3db529d5cc Revert "ar71xx: Add GRO support to ag71xx"
This reverts commit 13e5e47369.
This commit causes a severe regression in LAN->WAN routing performance
for several devices. This appears to be caused by the extra requirement
to validate the SKB checksum early in the rx path, which the ethernet
hardware does not do

Signed-off-by: Felix Fietkau <nbd@nbd.name>
2017-10-17 16:03:11 +02:00
Hauke Mehrtens
27f23bac8f ar71xx: fix section mismatches
Fix some section mismatches found with kernel 4.9.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2017-10-11 22:36:04 +02:00
Hauke Mehrtens
7bbf4117c6 ar71xx: Add kernel 4.9 support
This add support for kernel 4.9 to the ar71xx target.
It was compile tested with the generic, NAND and mikrotik subtarget.
Multiple members of the community tested it on their boards and did not
report any major problem so far.

Especially the NAND part received some changes to adapt to the new
kernel APIs. The serial driver hack used for the Arduino Yun was not
ported because the kernel changed there a lot.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2017-10-11 22:32:39 +02:00
Rosen Penev
13e5e47369 ar71xx: Add GRO support to ag71xx
On a TL-WN710N, this patch increases iperf performance from ~92.5 to ~93.5 mbps. Keep in mind the WN710N is a 100mbps device. I expect greater numbers from gigabit devices.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
2017-09-11 17:12:42 +02:00
Milan Krstić
4c1ce83548 ag71xx: add support for port mirroring
This exposes hardware port mirroring in ag71xx driver (e.g. TL-WR841ND) via
swconfig API.

Signed-off-by: Milan Krstić <milan.krstic@gmail.com>
2017-07-21 08:10:38 +02:00
Alexander Couzens
1025941662 ar71xx/ag71xx_ethtool: don't return uninitialized return value on success
ag71xx_ethtool_set_ringparam() will return an uninitialized value on
success.

Found-by: Coverity Scan #1330877
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
2017-06-11 14:22:44 +02:00
Alexander Couzens
64cd4b48e8 ar71xx/ag71xx_mdio_probe: fix a memory leak when probe fails
Found-by: Coverity Scan #1330233
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
2017-06-11 14:22:44 +02:00
Alexander Couzens
d18cb142d5 ar71xx/ag71xx_ar7240_get_port_link: fix off-by-one check on argument port
Found-by: Coverity Scan #1329901
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
2017-06-11 14:22:44 +02:00