* drivers/net/ethernet/apple
From: Geert Uytterhoeven @ 2011-10-25 20:19 UTC (permalink / raw)
To: Jeff Kirsher; +Cc: Benjamin Herrenschmidt, netdev, linux-kernel
Hi Jeff,
drivers/net/ethernet/apple/mac89x0.c is a driver for the Crystal Semiconductor
(Now Cirrus Logic) CS89[02]0, so it belongs in drivers/net/ethernet/cirrus,
next to cs89x0.c.
And according to drivers/net/ethernet/apple/mace.h, "mace" is the
"Am79C940 MACE (Medium Access Control for Ethernet)", so mace and
macmace should be in drivers/net/ethernet/amd/.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply
* Re: Linux 3.1-rc9
From: Simon Kirby @ 2011-10-25 20:20 UTC (permalink / raw)
To: Thomas Gleixner, David Miller
Cc: Peter Zijlstra, Linus Torvalds, Linux Kernel Mailing List,
Dave Jones, Martin Schwidefsky, Ingo Molnar, Network Development
In-Reply-To: <20111024190203.GA24410@hostway.ca>
On Mon, Oct 24, 2011 at 12:02:03PM -0700, Simon Kirby wrote:
> Ok, hit the hang about 4 more times, but only this morning on a box with
> a serial cable attached. Yay!
Here's lockdep output from another box. This one looks a bit different.
Simon-
[583223.799383] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[583223.805083] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[583223.805093]
[583223.805094] =================================
[583223.805096] [ INFO: inconsistent lock state ]
[583223.805098] 3.1.0-rc10-hw-lockdep+ #51
[583223.805100] ---------------------------------
[583223.805102] inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage.
[583223.805105] swapper/0 [HC0[0]:SC0[0]:HE1:SE1] takes:
[583223.805107] (slock-AF_INET){+.?.-.}, at: [<ffffffff81604540>] sk_clone+0x120/0x420
[583223.805116] {IN-SOFTIRQ-W} state was registered at:
[583223.805117] [<ffffffff81098c7c>] __lock_acquire+0xcbc/0x2180
[583223.805123] [<ffffffff8109a7b9>] lock_acquire+0x109/0x140
[583223.805126] [<ffffffff816f55fc>] _raw_spin_lock+0x3c/0x50
[583223.805131] [<ffffffff8166bd3d>] udp_queue_rcv_skb+0x26d/0x4b0
[583223.805135] [<ffffffff8166c6a3>] __udp4_lib_rcv+0x2f3/0x910
[583223.805139] [<ffffffff8166ccd5>] udp_rcv+0x15/0x20
[583223.805142] [<ffffffff81641960>] ip_local_deliver_finish+0x100/0x2f0
[583223.805146] [<ffffffff81641bdd>] ip_local_deliver+0x8d/0xa0
[583223.805149] [<ffffffff81641203>] ip_rcv_finish+0x1a3/0x510
[583223.805152] [<ffffffff816417e2>] ip_rcv+0x272/0x2f0
[583223.805154] [<ffffffff81610d67>] __netif_receive_skb+0x4d7/0x560
[583223.805158] [<ffffffff81610ec0>] process_backlog+0xd0/0x1e0
[583223.805161] [<ffffffff81613880>] net_rx_action+0x140/0x2c0
[583223.805164] [<ffffffff810640b8>] __do_softirq+0x138/0x250
[583223.805168] [<ffffffff817002bc>] call_softirq+0x1c/0x30
[583223.805172] [<ffffffff810153c5>] do_softirq+0x95/0xd0
[583223.805176] [<ffffffff81063ecd>] local_bh_enable+0xed/0x110
[583223.805179] [<ffffffff81614c48>] dev_queue_xmit+0x1a8/0x8a0
[583223.805181] [<ffffffff8161f1aa>] neigh_resolve_output+0x17a/0x220
[583223.805185] [<ffffffff81647d4c>] ip_finish_output+0x2ec/0x590
[583223.805188] [<ffffffff81648078>] ip_output+0x88/0xe0
[583223.805191] [<ffffffff81646cd8>] ip_local_out+0x28/0x80
[583223.805194] [<ffffffff81646d39>] ip_send_skb+0x9/0x40
[583223.805197] [<ffffffff8166aeb2>] udp_send_skb+0x122/0x390
[583223.805200] [<ffffffff8166db0c>] udp_sendmsg+0x7dc/0x920
[583223.805203] [<ffffffff81675e1f>] inet_sendmsg+0xbf/0x120
[583223.805207] [<ffffffff815ff333>] sock_sendmsg+0xe3/0x110
[583223.805209] [<ffffffff815ffc55>] sys_sendto+0x105/0x140
[583223.805212] [<ffffffff816fe052>] system_call_fastpath+0x16/0x1b
[583223.805217] irq event stamp: 4284605374
[583223.805219] hardirqs last enabled at (4284605372): [<ffffffff816101ad>] net_rps_action_and_irq_enable+0x8d/0xa0
[583223.805222] hardirqs last disabled at (4284605373): [<ffffffff8106412d>] __do_softirq+0x1ad/0x250
[583223.805226] softirqs last enabled at (4284605374): [<ffffffff81064176>] __do_softirq+0x1f6/0x250
[583223.805230] softirqs last disabled at (4284605313): [<ffffffff817002bc>] call_softirq+0x1c/0x30
[583223.805233]
[583223.805233] other info that might help us debug this:
[583223.805235] Possible unsafe locking scenario:
[583223.805236]
[583223.805237] CPU0
[583223.805238] ----
[583223.805239] lock(slock-AF_INET);
[583223.805241] <Interrupt>
[583223.805242] lock(slock-AF_INET);
[583223.805244]
[583223.805245] *** DEADLOCK ***
[583223.805246]
[583223.805248] 1 lock held by swapper/0:
[583223.805249] #0: (slock-AF_INET){+.?.-.}, at: [<ffffffff81604540>] sk_clone+0x120/0x420
[583223.805254]
[583223.805254] stack backtrace:
[583223.805257] Pid: 0, comm: swapper Not tainted 3.1.0-rc10-hw-lockdep+ #51
[583223.805259] Call Trace:
[583223.805264] [<ffffffff81096033>] print_usage_bug+0x243/0x310
[583223.805267] [<ffffffff810965b4>] mark_lock+0x4b4/0x6c0
[583223.805271] [<ffffffff81097400>] ? check_usage_forwards+0x110/0x110
[583223.805275] [<ffffffff81096862>] mark_held_locks+0xa2/0x130
[583223.805278] [<ffffffff816f6334>] ? retint_restore_args+0x13/0x13
[583223.805281] [<ffffffff81096c0d>] trace_hardirqs_on_caller+0x13d/0x1c0
[583223.805286] [<ffffffff813a60ae>] trace_hardirqs_on_thunk+0x3a/0x3f
[583223.805290] [<ffffffff81092b8e>] ? tick_nohz_stop_sched_tick+0x2fe/0x430
[583223.805293] [<ffffffff816f6334>] ? retint_restore_args+0x13/0x13
[583223.805297] [<ffffffff8101b80e>] ? mwait_idle+0x14e/0x170
[583223.805301] [<ffffffff8101b805>] ? mwait_idle+0x145/0x170
[583223.805304] [<ffffffff81013156>] cpu_idle+0x96/0xf0
[583223.805307] [<ffffffff816ca491>] rest_init+0xd1/0xe0
[583223.805310] [<ffffffff816ca3c0>] ? csum_partial_copy_generic+0x170/0x170
[583223.805315] [<ffffffff81adcc55>] start_kernel+0x360/0x3ac
[583223.805318] [<ffffffff81adc2a2>] x86_64_start_reservations+0x82/0x89
[583223.805321] [<ffffffff81adc3b8>] x86_64_start_kernel+0x10f/0x12a
[583223.805325] [<ffffffff81adc140>] ? early_idt_handlers+0x140/0x140
[583226.813848] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[583232.802948] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[583244.833571] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000101, exited with 00000102?
[583253.849631] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000101, exited with 00000102?
[583268.837126] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000101, exited with 00000102?
[587843.931805] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[587846.165584] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000101, exited with 00000102?
[587850.602316] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000101, exited with 00000102?
[587859.482841] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[587873.940136] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000102, exited with 00000103?
[587877.240624] huh, entered softirq 3 NET_RX ffffffff81613740 preempt_count 00000101, exited with 00000102?
[590476.272022] BUG: soft lockup - CPU#0 stuck for 22s! [swapper:0]
[590476.276002] Modules linked in: ipmi_devintf ipmi_si ipmi_msghandler xt_recent nf_conntrack_ftp xt_state xt_owner nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 bnx2
[590476.276002] irq event stamp: 4284605374
[590476.276002] hardirqs last enabled at (4284605372): [<ffffffff816101ad>] net_rps_action_and_irq_enable+0x8d/0xa0
[590476.276002] hardirqs last disabled at (4284605373): [<ffffffff8106412d>] __do_softirq+0x1ad/0x250
[590476.276002] softirqs last enabled at (4284605374): [<ffffffff81064176>] __do_softirq+0x1f6/0x250
[590476.276002] softirqs last disabled at (4284605313): [<ffffffff817002bc>] call_softirq+0x1c/0x30
[590476.276002] CPU 0
[590476.276002] Modules linked in: ipmi_devintf ipmi_si ipmi_msghandler xt_recent nf_conntrack_ftp xt_state xt_owner nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 bnx2
[590476.276002]
[590476.276002] Pid: 0, comm: swapper Not tainted 3.1.0-rc10-hw-lockdep+ #51 Dell Inc. PowerEdge 1950/0UR033
[590476.276002] RIP: 0010:[<ffffffff813a4ee3>] [<ffffffff813a4ee3>] delay_tsc+0x73/0xd0
[590476.276002] RSP: 0018:ffff88022fc03cf0 EFLAGS: 00000206
[590476.276002] RAX: 00042f884dcdaa24 RBX: ffff88022fc0d3c0 RCX: 000000004dcdaa24
[590476.380029] BUG: soft lockup - CPU#1 stuck for 22s! [php:10828]
[590476.380033] Modules linked in: ipmi_devintf ipmi_si ipmi_msghandler xt_recent nf_conntrack_ftp xt_state xt_owner nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 bnx2
[590476.380044] irq event stamp: 0
[590476.380045] hardirqs last enabled at (0): [< (null)>] (null)
[590476.380048] hardirqs last disabled at (0): [<ffffffff8105aa8b>] copy_process+0x65b/0x1450
[590476.380056] softirqs last enabled at (0): [<ffffffff8105aa8b>] copy_process+0x65b/0x1450
[590476.380060] softirqs last disabled at (0): [< (null)>] (null)
[590476.380063] CPU 1
[590476.380064] Modules linked in: ipmi_devintf ipmi_si ipmi_msghandler xt_recent nf_conntrack_ftp xt_state xt_owner nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 bnx2
[590476.380072]
[590476.380075] Pid: 10828, comm: php Not tainted 3.1.0-rc10-hw-lockdep+ #51 Dell Inc. PowerEdge 1950/0UR033
[590476.380079] RIP: 0010:[<ffffffff8101afa6>] [<ffffffff8101afa6>] native_read_tsc+0x6/0x20
[590476.380086] RSP: 0000:ffff88022fc43ce0 EFLAGS: 00000206
[590476.380088] RAX: 000000005aa56d04 RBX: ffffffff816f6334 RCX: 000000005aa56c92
[590476.380091] RDX: 0000000000042f88 RSI: ffffffff818f9896 RDI: 0000000000000001
[590476.380093] RBP: ffff88022fc43ce0 R08: 000000005aa56c92 R09: 0000000000000000
[590476.380096] R10: ffff88014b9a9f20 R11: 0000000000000000 R12: ffff88022fc43c58
[590476.380098] R13: ffffffff816feb33 R14: ffff88022fc43ce0 R15: 000000000e27878c
[590476.380101] FS: 00007fb61c8fa720(0000) GS:ffff88022fc40000(0000) knlGS:0000000000000000
[590476.380103] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[590476.380106] CR2: 00000000027914a0 CR3: 000000013a070000 CR4: 00000000000006e0
[590476.380108] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[590476.380110] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[590476.380113] Process php (pid: 10828, threadinfo ffff88014a1f2000, task ffff88014b9a9f20)
[590476.380115] Stack:
[590476.380117] ffff88022fc43d30 ffffffff813a4eaf ffff88014a1f2000 000000005aa56c38
[590476.380121] 00000001818f9896 ffff88001db58048 000000000e27878c 0000000076e96800
[590476.380125] 0000000000000001 ffff88014b9a9f20 ffff88022fc43d40 ffffffff813a4f6a
[590476.380129] Call Trace:
[590476.380132] <IRQ>
[590476.380137] [<ffffffff813a4eaf>] delay_tsc+0x3f/0xd0
[590476.380141] [<ffffffff813a4f6a>] __delay+0xa/0x10
[590476.380145] [<ffffffff813ac2ab>] do_raw_spin_lock+0x13b/0x180
[590476.380151] [<ffffffff816f5604>] _raw_spin_lock+0x44/0x50
[590476.380157] [<ffffffff81661823>] ? tcp_keepalive_timer+0x23/0x270
[590476.380161] [<ffffffff81661823>] tcp_keepalive_timer+0x23/0x270
[590476.380166] [<ffffffff8106cd5d>] run_timer_softirq+0x26d/0x410
[590476.380169] [<ffffffff8106cc88>] ? run_timer_softirq+0x198/0x410
[590476.380174] [<ffffffff81661800>] ? tcp_init_xmit_timers+0x20/0x20
[590476.380179] [<ffffffff810640b8>] __do_softirq+0x138/0x250
[590476.380184] [<ffffffff817002bc>] call_softirq+0x1c/0x30
[590476.380188] [<ffffffff810153c5>] do_softirq+0x95/0xd0
[590476.380191] [<ffffffff81063c8d>] irq_exit+0xdd/0x110
[590476.380196] [<ffffffff810310c9>] smp_apic_timer_interrupt+0x69/0xa0
[590476.380200] [<ffffffff816feb33>] apic_timer_interrupt+0x73/0x80
[590476.380203] <EOI>
[590476.380206] [<ffffffff816f6319>] ? retint_swapgs+0x13/0x1b
[590476.380208] Code: 08 ff 15 46 5c a1 00 48 c7 43 08 00 00 00 00 48 89 03 48 83 c4 08 5b c9 c3 66 66 66 2e 0f 1f 84 00 00 00 00 00 55 48 89 e5 0f 31
[590476.380227] c1 48 89 d0 48 c1 e0 20 89 ca 48 09 d0 c9 c3 66 2e 0f 1f 84
[590476.380236] Call Trace:
[590476.380237] <IRQ> [<ffffffff813a4eaf>] delay_tsc+0x3f/0xd0
[590476.380242] [<ffffffff813a4f6a>] __delay+0xa/0x10
[590476.380246] [<ffffffff813ac2ab>] do_raw_spin_lock+0x13b/0x180
[590476.380249] [<ffffffff816f5604>] _raw_spin_lock+0x44/0x50
[590476.380252] [<ffffffff81661823>] ? tcp_keepalive_timer+0x23/0x270
[590476.380256] [<ffffffff81661823>] tcp_keepalive_timer+0x23/0x270
[590476.380259] [<ffffffff8106cd5d>] run_timer_softirq+0x26d/0x410
[590476.380262] [<ffffffff8106cc88>] ? run_timer_softirq+0x198/0x410
[590476.380265] [<ffffffff81661800>] ? tcp_init_xmit_timers+0x20/0x20
[590476.380268] [<ffffffff810640b8>] __do_softirq+0x138/0x250
[590476.380271] [<ffffffff817002bc>] call_softirq+0x1c/0x30
[590476.380274] [<ffffffff810153c5>] do_softirq+0x95/0xd0
[590476.380277] [<ffffffff81063c8d>] irq_exit+0xdd/0x110
[590476.380280] [<ffffffff810310c9>] smp_apic_timer_interrupt+0x69/0xa0
[590476.380283] [<ffffffff816feb33>] apic_timer_interrupt+0x73/0x80
[590476.380285] <EOI> [<ffffffff816f6319>] ? retint_swapgs+0x13/0x1b
[590476.484032] BUG: soft lockup - CPU#2 stuck for 23s! [suexec:10831]
...
^ permalink raw reply
* Re: [PATCH v13 0/6] flexcan: Add support for powerpc flexcan (freescale p1010)
From: Scott Wood @ 2011-10-25 21:37 UTC (permalink / raw)
To: Kumar Gala
Cc: Robin Holt, netdev, U Bhaskar-B22300, socketcan-core, PPC list,
David S. Miller
In-Reply-To: <D79CB818-C14E-4C8D-9A8D-42B39ADE20B2@kernel.crashing.org>
On 10/18/2011 06:43 AM, Kumar Gala wrote:
>
>>> Robin,
>>>
>>> Do you remember why we went with just 'fsl,p1010-flexcan' as the device tree compatible? Do we feel the flex can on P1010 isn't the same as on MPC5xxx? or the ARM SoCs?
>>
>> The decision was due to the fact there is no true "generic" fsl.flexcan
>> chip free of any SOC implementation and therefore not something which
>> could be separately defined. That decision was made by Grant Likely.
>> I will inline that email below.
>>
>> Robin
>
>
> Thanks, I'll look into this internally at FSL. I think its confusing as hell to have "fsl,p1010-flexcan" in an ARM .dts
It's confusing to have devices labelled in vague ways that we can't tie
back to any real piece of hardware, or even a public architectural spec.
If you're talking to our hardware people, ask them to put public names
and versions, guaranteed unique throughout FSL, on all of our logic
blocks -- with public block manuals that have any SW-relevant
integration parameters clearly itemized.
Why is putting "fsl,p1010-flexcan" an an ARM device any more confusing
than putting it on some PowerPC chip that is not a p1010? Think of it
like a PCI ID, the actual value not being meaningful for much other than
its uniqueness and the ability to find a manual for the hardware.
This has been the recommended practice for quite some time.
> and don't think any reasonable ARM customer of FSL would know to put
> a PPC SOC name in their .dts.
If an ARM device tree comes along that just has
"fsl,some-arm-chip-flexcan", so what? Let the same driver bind against
both, again like PCI IDs. Additional compatibles are mainly a
convenience to give things a chance to work before the driver is updated
(a frequent irritant with PCI IDs and new hardware).
Ideally we would be publishing a sample device tree for our ARM chips
and their peripherals, though. :-P
> I'll ask the HW guys what's going on
> so we can come up with a bit more generic name so we don't have to
> constantly change this. Even if its just:
>
> fsl,ppc-flexcan & fsl,arm-flexcan.
Why is CPU instruction set relevant?
Would a QorIQ customer think to check for an existing compatible in
mpc5xxx, or even mpc83xx or mpc86xx?
-Scott
^ permalink raw reply
* unres_qlen and arp resolution/cache behavior
From: Gabe Black @ 2011-10-25 22:12 UTC (permalink / raw)
To: linux-kernel, netdev
Documentation about net.ipv4.neigh.*.unres_qlen seems to indicate that
it determines the number of frames that will be held while an arp is
being resolved.
Say packets start transmitting at a destination that needs to be (arp)
resolved. If the number of backlogged packets exceeds the queue
length, then the frames are dropped (from what I understand). My
question is what if the arp cached address expires and another arp is
issued. Does the same scenario occur (i.e. packets are backlogged
until that arp returns)? Or is the kernel smart enough to use the old
cached value until the arp returns, at which point it then uses the
updated address?
Thank you for your time,
Gabe
^ permalink raw reply
* [PATCH] net/ethernet: sc92031 is not Realtek
From: Cesar Eduardo Barros @ 2011-10-25 22:32 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, David S. Miller, Jeff Kirsher,
Realtek linux nic maintainers, Francois Romieu, Jeff Garzik,
Donald Becker, Cesar Eduardo Barros
While the SC92031 could be found on fake "Realtek" NICs, it has no
relationship to Realtek, and is actually from Silan.
Create a new subdirectory for silan and move sc92031 there.
Signed-off-by: Cesar Eduardo Barros <cesarb@cesarb.net>
---
drivers/net/ethernet/Kconfig | 1 +
drivers/net/ethernet/Makefile | 1 +
drivers/net/ethernet/realtek/Kconfig | 12 -------
drivers/net/ethernet/realtek/Makefile | 1 -
drivers/net/ethernet/silan/Kconfig | 33 +++++++++++++++++++++
drivers/net/ethernet/silan/Makefile | 5 +++
drivers/net/ethernet/{realtek => silan}/sc92031.c | 0
7 files changed, 40 insertions(+), 13 deletions(-)
create mode 100644 drivers/net/ethernet/silan/Kconfig
create mode 100644 drivers/net/ethernet/silan/Makefile
rename drivers/net/ethernet/{realtek => silan}/sc92031.c (100%)
diff --git a/drivers/net/ethernet/Kconfig b/drivers/net/ethernet/Kconfig
index 6dff5a0..597f4d4 100644
--- a/drivers/net/ethernet/Kconfig
+++ b/drivers/net/ethernet/Kconfig
@@ -159,6 +159,7 @@ config S6GMAC
will be called s6gmac.
source "drivers/net/ethernet/seeq/Kconfig"
+source "drivers/net/ethernet/silan/Kconfig"
source "drivers/net/ethernet/sis/Kconfig"
source "drivers/net/ethernet/sfc/Kconfig"
source "drivers/net/ethernet/sgi/Kconfig"
diff --git a/drivers/net/ethernet/Makefile b/drivers/net/ethernet/Makefile
index c53ad3a..be5dde0 100644
--- a/drivers/net/ethernet/Makefile
+++ b/drivers/net/ethernet/Makefile
@@ -58,6 +58,7 @@ obj-$(CONFIG_SH_ETH) += renesas/
obj-$(CONFIG_NET_VENDOR_RDC) += rdc/
obj-$(CONFIG_S6GMAC) += s6gmac.o
obj-$(CONFIG_NET_VENDOR_SEEQ) += seeq/
+obj-$(CONFIG_NET_VENDOR_SILAN) += silan/
obj-$(CONFIG_NET_VENDOR_SIS) += sis/
obj-$(CONFIG_SFC) += sfc/
obj-$(CONFIG_NET_VENDOR_SGI) += sgi/
diff --git a/drivers/net/ethernet/realtek/Kconfig b/drivers/net/ethernet/realtek/Kconfig
index 84083ec..0578859 100644
--- a/drivers/net/ethernet/realtek/Kconfig
+++ b/drivers/net/ethernet/realtek/Kconfig
@@ -115,16 +115,4 @@ config R8169
To compile this driver as a module, choose M here: the module
will be called r8169. This is recommended.
-config SC92031
- tristate "Silan SC92031 PCI Fast Ethernet Adapter driver (EXPERIMENTAL)"
- depends on PCI && EXPERIMENTAL
- select CRC32
- ---help---
- This is a driver for the Fast Ethernet PCI network cards based on
- the Silan SC92031 chip (sometimes also called Rsltek 8139D). If you
- have one of these, say Y here.
-
- To compile this driver as a module, choose M here: the module
- will be called sc92031. This is recommended.
-
endif # NET_VENDOR_REALTEK
diff --git a/drivers/net/ethernet/realtek/Makefile b/drivers/net/ethernet/realtek/Makefile
index e48cfb6..71b1da3 100644
--- a/drivers/net/ethernet/realtek/Makefile
+++ b/drivers/net/ethernet/realtek/Makefile
@@ -6,4 +6,3 @@ obj-$(CONFIG_8139CP) += 8139cp.o
obj-$(CONFIG_8139TOO) += 8139too.o
obj-$(CONFIG_ATP) += atp.o
obj-$(CONFIG_R8169) += r8169.o
-obj-$(CONFIG_SC92031) += sc92031.o
diff --git a/drivers/net/ethernet/silan/Kconfig b/drivers/net/ethernet/silan/Kconfig
new file mode 100644
index 0000000..ae1ce17
--- /dev/null
+++ b/drivers/net/ethernet/silan/Kconfig
@@ -0,0 +1,33 @@
+#
+# Silan device configuration
+#
+
+config NET_VENDOR_SILAN
+ bool "Silan devices"
+ default y
+ depends on PCI && EXPERIMENTAL
+ ---help---
+ If you have a network (Ethernet) card belonging to this class, say Y
+ and read the Ethernet-HOWTO, available from
+ <http://www.tldp.org/docs.html#howto>.
+
+ Note that the answer to this question doesn't directly affect the
+ kernel: saying N will just cause the configurator to skip all
+ the questions about Silan devices. If you say Y, you will be asked for
+ your specific card in the following questions.
+
+if NET_VENDOR_SILAN
+
+config SC92031
+ tristate "Silan SC92031 PCI Fast Ethernet Adapter driver (EXPERIMENTAL)"
+ depends on PCI && EXPERIMENTAL
+ select CRC32
+ ---help---
+ This is a driver for the Fast Ethernet PCI network cards based on
+ the Silan SC92031 chip (sometimes also called Rsltek 8139D). If you
+ have one of these, say Y here.
+
+ To compile this driver as a module, choose M here: the module
+ will be called sc92031. This is recommended.
+
+endif # NET_VENDOR_SILAN
diff --git a/drivers/net/ethernet/silan/Makefile b/drivers/net/ethernet/silan/Makefile
new file mode 100644
index 0000000..4ad3523
--- /dev/null
+++ b/drivers/net/ethernet/silan/Makefile
@@ -0,0 +1,5 @@
+#
+# Makefile for the Silan network device drivers.
+#
+
+obj-$(CONFIG_SC92031) += sc92031.o
diff --git a/drivers/net/ethernet/realtek/sc92031.c b/drivers/net/ethernet/silan/sc92031.c
similarity index 100%
rename from drivers/net/ethernet/realtek/sc92031.c
rename to drivers/net/ethernet/silan/sc92031.c
--
1.7.6.4
^ permalink raw reply related
* Re: Linux 3.1-rc9
From: David Miller @ 2011-10-25 23:18 UTC (permalink / raw)
To: tglx
Cc: torvalds, sim, netdev, a.p.zijlstra, linux-kernel, davej,
schwidefsky, mingo
In-Reply-To: <alpine.LFD.2.02.1110251419170.21837@ionos>
From: Thomas Gleixner <tglx@linutronix.de>
Date: Tue, 25 Oct 2011 14:30:50 +0200 (CEST)
> Subject: net: Unlock sock before calling sk_free()
>
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Good spotting, applied, thanks Thomas!
^ permalink raw reply
* BUSINESS PROPOSAL, PLEASE CONTACT ME
From: MR RAYMOND CHOONG YEE @ 2011-10-25 23:16 UTC (permalink / raw)
Hello, I am Mr.Raymond Choong Yee H, i seek your attention for the confirmation of a Project. I shall provide you with more details after hearing from you.contact me via (raymondchoongyh@9.cn)
^ permalink raw reply
* Re: [PATCH] net: make bonding slaves honour master's skb->priority
From: David Miller @ 2011-10-25 23:22 UTC (permalink / raw)
To: fubar; +Cc: zenczykowski, fbl, netdev
In-Reply-To: <7147.1319561135@death>
From: Jay Vosburgh <fubar@us.ibm.com>
Date: Tue, 25 Oct 2011 09:45:35 -0700
> Maciej Żenczykowski <zenczykowski@gmail.com> wrote:
>
>>> I tripped on it few times and never understood why we
>>> were fixing the priority at this point.
>>> thanks,
>>
>>It looks like ancient code, going all the way back to something like
>>Linux 2.4.14 (if I recall my git blame session correctly).
>>I can't think of a good reason for this...
>
> I believe this line goes back to the beginning, prior to git or
> bk. The skb->priority probably meant something different than it does
> today.
>
> In any event, I'm fine with the code change. There's still no
> actual log message, but if Davem is ok with that I'm ok with it, given
> the trivial nature of the change.
>
> Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Applied, thanks everyone.
^ permalink raw reply
* Re: [PATCH] Fix caif BUG() with network namespaces
From: David Miller @ 2011-10-25 23:22 UTC (permalink / raw)
To: dwmw2; +Cc: sjurbren, netdev
In-Reply-To: <1319527522.24797.10.camel@shinybook.infradead.org>
From: David Woodhouse <dwmw2@infradead.org>
Date: Tue, 25 Oct 2011 09:25:21 +0200
> The caif code will register its own pernet_operations, and then register
> a netdevice_notifier. Each time the netdevice_notifier is triggered,
> it'll do some stuff... including a lookup of its own pernet stuff with
> net_generic().
>
> If the net_generic() call ever returns NULL, the caif code will BUG().
> That doesn't seem *so* unreasonable, I suppose — it does seem like it
> should never happen.
>
> However, it *does* happen. When we clone a network namespace,
> setup_net() runs through all the pernet_operations one at a time. It
> gets to loopback before it gets to caif. And loopback_net_init()
> registers a netdevice... while caif hasn't been initialised. So the caif
> netdevice notifier triggers, and immediately goes BUG().
>
> We could imagine a complex and overengineered solution to this generic
> class of problems, but this patch takes the simple approach. It just
> makes caif_device_notify() *not* go looking for its pernet data
> structures if the device it's being notified about isn't a caif device
> in the first place.
>
> Cc: stable@kernel.org
> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
> Acked-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Applied, thanks.
^ permalink raw reply
* BUSINESS PROPOSAL, PLEASE CONTACT ME
From: MR RAYMOND CHOONG YEE @ 2011-10-25 23:09 UTC (permalink / raw)
Hello, I am Mr.Raymond Choong Yee H, i seek your attention for the confirmation of a Project. I shall provide you with more details after hearing from you.contact me via (raymondchoongyh@9.cn)
^ permalink raw reply
* Re: [PATCH net-next] ipv4: use IS_ENABLED() macro to cleanup code
From: David Miller @ 2011-10-25 23:30 UTC (permalink / raw)
To: eric.dumazet; +Cc: netdev
In-Reply-To: <1319526752.3834.12.camel@edumazet-laptop>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Tue, 25 Oct 2011 09:12:32 +0200
> Remove some ugly #ifdef, using IS_ENABLED(CONFIG_xxx) macro
>
> IS_ENABLED(CONFIG_xxx) evaluates to 1 if CONFIG_xxx is set to 'y' or 'm'
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
I can understand the desire to do this, but it is going to run
afoul in cases where the definitions of datastructures in header
files are protected by real CPP ifdef guards rather than these
run-time variants.
So you build tested this with IPV6 disabled, right?
Didn't think so...
net/ipv4/ip_sockglue.c: In function ‘do_ip_setsockopt’:
net/ipv4/ip_sockglue.c:523:29: error: ‘LOOPBACK4_IPV6’ undeclared (first use in this function)
net/ipv4/ip_sockglue.c:523:29: note: each undeclared identifier is reported only once for each function it appears in
This fails because ip_sockglue.c guards the net/transp_v6.h header
inclusion with a real CPP guard.
^ permalink raw reply
* Re: Bug#645308: tg3 broken for NetXtreme 5714S in squeeze 6.0.3 installer
From: Matt Carlson @ 2011-10-26 0:20 UTC (permalink / raw)
To: Ben Hutchings
Cc: Matthew Carlson, Michael Chan, 645308@bugs.debian.org, Marc Haber,
netdev
In-Reply-To: <1319500074.31243.47.camel@deadeye>
On Mon, Oct 24, 2011 at 04:47:54PM -0700, Ben Hutchings wrote:
> On Mon, 2011-10-24 at 14:24 -0700, Matt Carlson wrote:
> > On Fri, Oct 21, 2011 at 05:19:39AM -0700, Ben Hutchings wrote:
> > > On Fri, 2011-10-21 at 11:08 +0200, Marc Haber wrote:
> > > > On Fri, Oct 21, 2011 at 11:00:46AM +0200, Marc Haber wrote:
> > > > > On Thu, Oct 20, 2011 at 05:28:34AM +0100, Ben Hutchings wrote:
> > > > > > I don't see any changes that would obviously change the way this device
> > > > > > is reconfigured during a down/up cycle. There were some changes to
> > > > > > power management that should just let the PCI core do some work that the
> > > > > > driver used to, but it's possible that the result isn't quite the same.
> > > > > > I built a module with those reverted; source and binary attached. Could
> > > > > > you test that? I checked that d-i does include an insmod command.
> > > > >
> > > > > The squeeze 6.0.3 installer with the shipped tg3.ko replaced with
> > > > > yours boots and networks just fine without any workaround and without
> > > > > manual interaction.
> > > >
> > > > I was a bit fast on that. The interface now fails right in the middle
> > > > of installation and needs the modprobe -r, modprobe stunt to network
> > > > again.
> > >
> > > Matt, Michael,
> > >
> > > The tg3 driver has regressed for the 5714S since Linux 2.6.32. Marc
> > > Haber found this in the backported version included in our stable
> > > update, but also confirmed it in Linux 3.0.
> > >
> > > Bringing the interface down and then up again (which the installer does
> > > for some reason) can leave it unable to pass traffic (possibly after
> > > working for a few packets) until the module is reloaded.
> > >
> > > I asked Marc to check whether reverting the power management changes
> > > (071697e2bcd8dff2af4d6fdd6525c2324f89553b,
> > > d237d9ecf06a00f0ebca657958cf2a1e92940796) made a difference, but it
> > > doesn't seem to.
> > >
> > > There is more information in the bug log at
> > > <http://bugs.debian.org/645308>.
> >
> > Where can I get the sources for this driver? Commit
> > 9e975cc291d80d5e4562d6bed15ec171e896d69b, entitled
> > "tg3: Fix io failures after chip reset" has been a common source of
> > problems.
>
> Our current package has Linux 3.0.6 which includes the backport of that
> change. However, it is *not* included in my backport to 2.6.32 so it
> doesn't explain the original report.
>
> The backported version can be found in:
>
> git://anonscm.debian.org/kernel/linux-2.6.git squeeze
The kernel version of that repository is 3.0.0-rc1. Am I looking in the
right place?
But you're right. The version of the driver in that repository does not
have the change.
^ permalink raw reply
* [GIT] Networking
From: David Miller @ 2011-10-26 0:32 UTC (permalink / raw)
To: torvalds; +Cc: akpm, netdev, linux-kernel
Three bug fixes here, including the sk_clone() error path unlock
fix from Thomas Gleixner.
As for the wireless warning spew, it is my understanding that person
most able to fix that bug is in the same room with you right now (hi
Johannes Berg :-)
Please pull, thanks a lot!
The following changes since commit 7670c7010c7b367ca40c3aba02afb36149764a6e:
Merge branch 'for-linus' of git://github.com/gregungerer/m68knommu (2011-10-25 16:25:48 +0200)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/davem/net.git master
David Woodhouse (1):
caif: Fix BUG() with network namespaces
Maciej Żenczykowski (1):
net: make bonding slaves honour master's skb->priority
Thomas Gleixner (1):
net: Unlock sock before calling sk_free()
drivers/net/bonding/bond_main.c | 1 -
net/caif/caif_dev.c | 5 +++--
net/core/sock.c | 1 +
3 files changed, 4 insertions(+), 3 deletions(-)
^ permalink raw reply
* erh你好1
From: erh @ 2011-10-26 1:52 UTC (permalink / raw)
本公司现有(商品销售；建筑工程；装修设计；广告；租赁；服务咨询;加工修理；运输；餐饮住宿;
海关;增值税等等 )。如贵公司需要我司代开發票，需提前向我司提供准确详细的资料清单；如贵公
司所在地是其它省、市的，本公司都是以快递投递的方式 (24小时至48小时到达)，并可上网查询验
证。收到确认无误后再给我司划账(数量金额大可再商讨优惠税率)。详情欢迎来电咨询! 顺祝商祺.
联系电话13692264168周益胜QQ1979969374
^ permalink raw reply
* [PATCH] Add TCP_NO_DELAYED_ACK socket option
From: Andy Lutomirski @ 2011-10-26 2:25 UTC (permalink / raw)
To: netdev; +Cc: Andy Lutomirski
When talking to an unfixable interactive peer that fails to set
TCP_NODELAY, disabling delayed ACKs can help mitigate the problem.
This is an evil thing to do, but if the entire network is private,
it's not that evil.
This works around a problem with the remote *application*, so make
it a socket option instead of a sysctl or a per-route option.
Signed-off-by: Andy Lutomirski <luto@amacapital.net>
---
This patch is a bit embarrassing. We talk to remote applications over
TCP that are very much interactive but don't set TCP_NODELAY. These
applications apparently cannot be fixed. As a partial workaround, if we
ACK every incoming segment, then as long as they don't transmit two
segments per rtt, we do pretty well.
Windows can do something similar, but it's per interface instead of per
socket:
http://support.microsoft.com/kb/328890
include/linux/tcp.h | 1 +
include/net/inet_connection_sock.h | 3 ++-
net/ipv4/tcp.c | 11 +++++++++++
net/ipv4/tcp_input.c | 3 ++-
4 files changed, 16 insertions(+), 2 deletions(-)
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 531ede8..2116f31 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -106,6 +106,7 @@ enum {
#define TCP_THIN_LINEAR_TIMEOUTS 16 /* Use linear timeouts for thin streams*/
#define TCP_THIN_DUPACK 17 /* Fast retrans. after 1 dupack */
#define TCP_USER_TIMEOUT 18 /* How long for loss retry before timeout */
+#define TCP_NO_DELAYED_ACK 19 /* Do not delay ACKs. */
/* for TCP_INFO socket option */
#define TCPI_OPT_TIMESTAMPS 1
diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h
index e6db62e..1ad91bf 100644
--- a/include/net/inet_connection_sock.h
+++ b/include/net/inet_connection_sock.h
@@ -106,8 +106,9 @@ struct inet_connection_sock {
struct {
__u8 pending; /* ACK is pending */
__u8 quick; /* Scheduled number of quick acks */
- __u8 pingpong; /* The session is interactive */
__u8 blocked; /* Delayed ACK was blocked by socket lock */
+ __u8 pingpong:1; /* The session is interactive */
+ __u8 nodelack:1; /* Delayed ACKs are disabled */
__u32 ato; /* Predicted tick of soft clock */
unsigned long timeout; /* Currently scheduled timeout */
__u32 lrcvtime; /* timestamp of last received data packet */
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 46febca..e8e98dc 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2385,6 +2385,13 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
}
break;
+ case TCP_NO_DELAYED_ACK:
+ if (val == 0 || val == 1)
+ icsk->icsk_ack.nodelack = !!val;
+ else
+ err = -EINVAL;
+ break;
+
#ifdef CONFIG_TCP_MD5SIG
case TCP_MD5SIG:
/* Read the IP->Key mappings from userspace */
@@ -2564,6 +2571,10 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
val = !icsk->icsk_ack.pingpong;
break;
+ case TCP_NO_DELAYED_ACK:
+ val = icsk->icsk_ack.nodelack;
+ break;
+
case TCP_CONGESTION:
if (get_user(len, optlen))
return -EFAULT;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 21fab3e..e7d7ee0 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -197,7 +197,8 @@ static void tcp_enter_quickack_mode(struct sock *sk)
static inline int tcp_in_quickack_mode(const struct sock *sk)
{
const struct inet_connection_sock *icsk = inet_csk(sk);
- return icsk->icsk_ack.quick && !icsk->icsk_ack.pingpong;
+ return (icsk->icsk_ack.quick && !icsk->icsk_ack.pingpong) ||
+ icsk->icsk_ack.nodelack;
}
static inline void TCP_ECN_queue_cwr(struct tcp_sock *tp)
--
1.7.6.4
^ permalink raw reply related
* Re: PROBLEM: System call 'sendmsg' of process ospfd (quagga) causes kernel oops
From: David Miller @ 2011-10-26 3:12 UTC (permalink / raw)
To: herbert; +Cc: eric.dumazet, evonlanthen, linux-kernel, netdev, timo.teras
In-Reply-To: <20111025115425.GA1530@gondor.apana.org.au>
From: Herbert Xu <herbert@gondor.hengli.com.au>
Date: Tue, 25 Oct 2011 13:54:25 +0200
> So I'm going to get rid of LL_ALLOCATED_SPACE completely and
> replace it with explicit references to the tailroom as it doesn't
> need the alignment anyway (The headroom needs alignment since
> we use it to ensure the head is aligned).
Ok.
^ permalink raw reply
* Re: [PATCH V2 2/4] MIPS: Add board support for Loongson1B
From: Kelvin Cheung @ 2011-10-26 4:27 UTC (permalink / raw)
To: Giuseppe CAVALLARO
Cc: Wu Zhangjin, linux-mips, linux-kernel, ralf, r0bertz, netdev
In-Reply-To: <4EA6716A.80307@st.com>
[-- Attachment #1: Type: text/plain, Size: 4272 bytes --]
Hi Giuseppe,
This patch works well on Loongson1B platform except one thing.
The rx checksum offload of normal descriptor is disabled by default.
So, I enabled this functon. And one minor tweak is added to your patch.
What about your opinion?
BTW, tx checksum insertion works now.
2011/10/25, Giuseppe CAVALLARO <peppe.cavallaro@st.com>:
> On 10/25/2011 9:09 AM, Giuseppe CAVALLARO wrote:
>> On 10/25/2011 4:12 AM, Kelvin Cheung wrote:
>>> 2011/10/24, Giuseppe CAVALLARO <peppe.cavallaro@st.com>:
>>>> On 10/24/2011 4:05 PM, Kelvin Cheung wrote:
>>>>> 2011/10/24, Giuseppe CAVALLARO <peppe.cavallaro@st.com>:
>>>>>> Hello Kelvin.
>>>>>>
>>>>>> On 10/24/2011 12:36 PM, Kelvin Cheung wrote:
>>>>>>
>>>>>> [snip]
>>>>>>
>>>>>>> According to datasheet of Loongson 1B, the buffer size in RX/TX
>>>>>>> descriptor is only 2KB. So the Loongson1B's GMAC could not handle
>>>>>>> jumbo frames. And the second buffer is useless in this case. Am I
>>>>>>> right? Is there a better way than ifdef CONFIG_MACH_LOONGSON1 to
>>>>>>> avoid duplicate code?
>>>>>>
>>>>>> Sorry for my misunderstanding.
>>>>>>
>>>>>> I think you have to use the normal descriptor and remove the enh_desc
>>>>>> from the platform w/o modifying the driver at all.
>>>>>>
>>>>>> The driver will be able to select/configure all automatically (also
>>>>>> jumbo).
>>>>>>
>>>>>> Let me know.
>>>>>
>>>>> That's the problem.
>>>>> The bitfield definition of Loongson1B is also different from normal
>>>>> descriptor.
>>>>
>>>> The problem is not in the Loongson1B gmac.
>>>
>>> I found that the bit checksum_insertion is not existed in normal
>>> descriptor.
>>>
>>>> The normal descriptor fields in the stmmac refer to an old synopsys
>>>> databook.
>>>
>>> Could you send me the new databook of Synopsys GMAC?
>>>
>>>> New chips have the same structure you have added; so we should fix this
>>>> in the driver w/o breaking the compatibility for old chips.
>>>
>>> Agree.
>>>
>>>> I kindly ask you to confirm if the currently normal descriptor structure
>>>> (w/o your changes) doesn't work on your platform.
>>>> Did you test it?
>>>
>>> Well, the normal descriptor works on my platform except TX checksum
>>> offload.
>>
>> ok! I suspected that.
>>
>>
>>>>> Moreover, I want to enable the TX checksum offload function which is
>>>>> not supported in normal descriptor.
>>>>> Any suggestions?
>>>>
>>>> It is supported but you have to pass from the platform: tx_coe = 1.
>>>
>>> I noticed that the flag csum_insertion is passed to
>>> ndesc_prepare_tx_desc() in stmmac_xmit(). But ndesc_prepare_tx_desc()
>>> just ignores it.
>>> In other words, the TX checksum offload function is disabled in normal
>>> descriptor currently.
>>>
>>> Should we fix this problem for normal descriptor?
>>
>> Yes, we should. If you agree, I'll update the normal descriptor
>> structure to yours. This is the normal descriptor used in newer GMAC.
>> Tx csum will be done for normal descriptors in case of these GMAC
>> devices and not for old MAC10/100. For the MAC10/100 some bits for
>> normal descriptors are reserved and won't be used at all.
>>
>> I'll also verify that the patch doesn't break the back-compatibility
>> with old MAC10/100. I have the HW where doing the tests.
>>
>> After that, I'll prepare the patch for net-next and for your kernel.
>
> Hello Kelvin
>
> attached the patch tested on my development kernel.
> It runs fine on old and new mac devices.
>
> Can you try it on your side? Hmm, it is likely it won't apply fine on
> your tree but you know the changes ;-).
>
> If ok, I'll rework it for net-next and send it to the mailing list.
>
> Thanks
> Peppe
>
>>
>>>
>>>> Peppe
>>>>>
>>>>>> Note:
>>>>>> IIRC, there is a bit difference in case of normal descriptors for
>>>>>> Synopsys databook newer than the 1.91 (I used for testing this mode).
>>>>>> In any case, I remember that, on some platforms, the normal
>>>>>> descriptors
>>>>>> have been used w/o problems also on these new chip generations.
>>>>>>
>>>>>> Peppe
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>> --
>> 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
>>
>
>
--
Best Regards!
Kelvin
[-- Attachment #2: 0001-stmmac-update-normal-descriptor-structure.patch --]
[-- Type: text/x-patch, Size: 7719 bytes --]
From 93ee3effbde61013a886d579fc89c36ddddb35a4 Mon Sep 17 00:00:00 2001
From: Kelvin Cheung <keguang.zhang@gmail.com>
Date: Wed, 26 Oct 2011 11:08:01 +0800
Subject: [PATCH] stmmac: update normal descriptor structure
This patch updates the normal descriptor structure
to work fine on new GMAC Synopsys chips.
Normal descriptors were designed on the old MAC10/100
databook 1.91 where some bits were reserved: for example
the tx checksum insertion and rx checksum offload.
The patch maintains the back-compatibility with old
MAC devices (tested on STx7109 MAC10/100) and adds new
fields that actually new GMAC devices can use.
For example, STx7109 will pass from the platform
tx_coe = 0, enh_desc = 0, has_gmac = 0.
A platform like Loongson1B (Mips) will pass:
tx_coe = 1, enh_desc = 0, has_gmac = 1.
Thanks to Kelvin, he enhanced the normal descriptors for
GMAC on Loongson1B.
Signed-off-by: Kelvin Cheung <keguang.zhang@gmail.com>
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/stmmac/common.h | 8 +++---
drivers/net/stmmac/descs.h | 31 +++++++++++++++-----------
drivers/net/stmmac/norm_desc.c | 40 +++++++++++++++++++---------------
drivers/net/stmmac/stmmac_ethtool.c | 8 +++---
4 files changed, 48 insertions(+), 39 deletions(-)
diff --git a/drivers/net/stmmac/common.h b/drivers/net/stmmac/common.h
index 375ea19..6ec4a8c 100644
--- a/drivers/net/stmmac/common.h
+++ b/drivers/net/stmmac/common.h
@@ -48,7 +48,7 @@ struct stmmac_extra_stats {
unsigned long tx_underflow ____cacheline_aligned;
unsigned long tx_carrier;
unsigned long tx_losscarrier;
- unsigned long tx_heartbeat;
+ unsigned long vlan_tag;
unsigned long tx_deferred;
unsigned long tx_vlan;
unsigned long tx_jabber;
@@ -57,9 +57,9 @@ struct stmmac_extra_stats {
unsigned long tx_ip_header_error;
/* Receive errors */
unsigned long rx_desc;
- unsigned long rx_partial;
- unsigned long rx_runt;
- unsigned long rx_toolong;
+ unsigned long sa_filter_fail;
+ unsigned long overflow_error;
+ unsigned long ipc_csum_error;
unsigned long rx_collision;
unsigned long rx_crc;
unsigned long rx_length;
diff --git a/drivers/net/stmmac/descs.h b/drivers/net/stmmac/descs.h
index 63a03e2..9820ec8 100644
--- a/drivers/net/stmmac/descs.h
+++ b/drivers/net/stmmac/descs.h
@@ -25,33 +25,34 @@ struct dma_desc {
union {
struct {
/* RDES0 */
- u32 reserved1:1;
+ u32 payload_csum_error:1;
u32 crc_error:1;
u32 dribbling:1;
u32 mii_error:1;
u32 receive_watchdog:1;
u32 frame_type:1;
u32 collision:1;
- u32 frame_too_long:1;
+ u32 ipc_csum_error:1;
u32 last_descriptor:1;
u32 first_descriptor:1;
- u32 multicast_frame:1;
- u32 run_frame:1;
+ u32 vlan_tag:1;
+ u32 overflow_error:1;
u32 length_error:1;
- u32 partial_frame_error:1;
+ u32 sa_filter_fail:1;
u32 descriptor_error:1;
u32 error_summary:1;
u32 frame_length:14;
- u32 filtering_fail:1;
+ u32 da_filter_fail:1;
u32 own:1;
/* RDES1 */
u32 buffer1_size:11;
u32 buffer2_size:11;
- u32 reserved2:2;
+ u32 reserved1:2;
u32 second_address_chained:1;
u32 end_ring:1;
- u32 reserved3:5;
+ u32 reserved2:5;
u32 disable_ic:1;
+
} rx;
struct {
/* RDES0 */
@@ -91,24 +92,28 @@ struct dma_desc {
u32 underflow_error:1;
u32 excessive_deferral:1;
u32 collision_count:4;
- u32 heartbeat_fail:1;
+ u32 vlan_frame:1;
u32 excessive_collisions:1;
u32 late_collision:1;
u32 no_carrier:1;
u32 loss_carrier:1;
- u32 reserved1:3;
+ u32 payload_error:1;
+ u32 frame_flushed:1;
+ u32 jabber_timeout:1;
u32 error_summary:1;
- u32 reserved2:15;
+ u32 ip_header_error:1;
+ u32 time_stamp_status:1;
+ u32 reserved1:13;
u32 own:1;
/* TDES1 */
u32 buffer1_size:11;
u32 buffer2_size:11;
- u32 reserved3:1;
+ u32 time_stamp_enable:1;
u32 disable_padding:1;
u32 second_address_chained:1;
u32 end_ring:1;
u32 crc_disable:1;
- u32 reserved4:2;
+ u32 checksum_insertion:2;
u32 first_segment:1;
u32 last_segment:1;
u32 interrupt:1;
diff --git a/drivers/net/stmmac/norm_desc.c b/drivers/net/stmmac/norm_desc.c
index 7008c29..fab8f0c 100644
--- a/drivers/net/stmmac/norm_desc.c
+++ b/drivers/net/stmmac/norm_desc.c
@@ -49,11 +49,12 @@ static int ndesc_get_tx_status(void *data, struct stmmac_extra_stats *x,
stats->collisions += p->des01.tx.collision_count;
ret = -1;
}
- if (unlikely(p->des01.tx.heartbeat_fail)) {
- x->tx_heartbeat++;
- stats->tx_heartbeat_errors++;
- ret = -1;
- }
+
+ if (p->des01.etx.vlan_frame) {
+ CHIP_DBG(KERN_INFO "GMAC TX status: VLAN frame\n");
+ x->tx_vlan++;
+ }
+
if (unlikely(p->des01.tx.deferred))
x->tx_deferred++;
@@ -66,13 +67,11 @@ static int ndesc_get_tx_len(struct dma_desc *p)
}
/* This function verifies if each incoming frame has some errors
- * and, if required, updates the multicast statistics.
- * In case of success, it returns csum_none because the device
- * is not able to compute the csum in HW. */
+ * and, if required, updates the multicast statistics. */
static int ndesc_get_rx_status(void *data, struct stmmac_extra_stats *x,
struct dma_desc *p)
{
- int ret = csum_none;
+ int ret = good_frame;
struct net_device_stats *stats = (struct net_device_stats *)data;
if (unlikely(p->des01.rx.last_descriptor == 0)) {
@@ -85,12 +84,12 @@ static int ndesc_get_rx_status(void *data, struct stmmac_extra_stats *x,
if (unlikely(p->des01.rx.error_summary)) {
if (unlikely(p->des01.rx.descriptor_error))
x->rx_desc++;
- if (unlikely(p->des01.rx.partial_frame_error))
- x->rx_partial++;
- if (unlikely(p->des01.rx.run_frame))
- x->rx_runt++;
- if (unlikely(p->des01.rx.frame_too_long))
- x->rx_toolong++;
+ if (unlikely(p->des01.rx.sa_filter_fail))
+ x->sa_filter_fail++;
+ if (unlikely(p->des01.rx.overflow_error))
+ x->overflow_error++;
+ if (unlikely(p->des01.rx.ipc_csum_error))
+ x->ipc_csum_error++;
if (unlikely(p->des01.rx.collision)) {
x->rx_collision++;
stats->collisions++;
@@ -112,10 +111,12 @@ static int ndesc_get_rx_status(void *data, struct stmmac_extra_stats *x,
x->rx_mii++;
ret = discard_frame;
}
- if (p->des01.rx.multicast_frame) {
- x->rx_multicast++;
- stats->multicast++;
+#ifdef STMMAC_VLAN_TAG_USED
+ if (p->des01.rx.vlan_tag) {
+ x->vlan_tag++;
+ stats->vlan_tag++;
}
+#endif
return ret;
}
@@ -184,6 +185,9 @@ static void ndesc_prepare_tx_desc(struct dma_desc *p, int is_fs, int len,
{
p->des01.tx.first_segment = is_fs;
norm_set_tx_desc_len(p, len);
+
+ if (likely(csum_flag))
+ p->des01.tx.checksum_insertion = cic_full;
}
static void ndesc_clear_tx_ic(struct dma_desc *p)
diff --git a/drivers/net/stmmac/stmmac_ethtool.c b/drivers/net/stmmac/stmmac_ethtool.c
index 7ed8fb6..3d58fcc 100644
--- a/drivers/net/stmmac/stmmac_ethtool.c
+++ b/drivers/net/stmmac/stmmac_ethtool.c
@@ -50,7 +50,7 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(tx_underflow),
STMMAC_STAT(tx_carrier),
STMMAC_STAT(tx_losscarrier),
- STMMAC_STAT(tx_heartbeat),
+ STMMAC_STAT(vlan_tag),
STMMAC_STAT(tx_deferred),
STMMAC_STAT(tx_vlan),
STMMAC_STAT(rx_vlan),
@@ -59,9 +59,9 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(tx_payload_error),
STMMAC_STAT(tx_ip_header_error),
STMMAC_STAT(rx_desc),
- STMMAC_STAT(rx_partial),
- STMMAC_STAT(rx_runt),
- STMMAC_STAT(rx_toolong),
+ STMMAC_STAT(sa_filter_fail),
+ STMMAC_STAT(overflow_error),
+ STMMAC_STAT(ipc_csum_error),
STMMAC_STAT(rx_collision),
STMMAC_STAT(rx_crc),
STMMAC_STAT(rx_length),
--
1.7.1
^ permalink raw reply related
* Re: [RFC v2 PATCH 5/4 PATCH] virtio-net: send gratuitous packet when needed
From: Jason Wang @ 2011-10-26 4:49 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: aliguori, kvm, quintela, jan.kiszka, Rusty Russell, qemu-devel,
blauwirbel, netdev, pbonzini
In-Reply-To: <20111025154106.GA22553@redhat.com>
On 10/25/2011 11:41 PM, Michael S. Tsirkin wrote:
> On Tue, Oct 25, 2011 at 10:50:41AM +0800, Jason Wang wrote:
>> On 10/24/2011 01:25 PM, Michael S. Tsirkin wrote:
>>> On Mon, Oct 24, 2011 at 02:54:59PM +1030, Rusty Russell wrote:
>>>> On Sat, 22 Oct 2011 13:43:11 +0800, Jason Wang <jasowang@redhat.com> wrote:
>>>>> This make let virtio-net driver can send gratituous packet by a new
>>>>> config bit - VIRTIO_NET_S_ANNOUNCE in each config update
>>>>> interrupt. When this bit is set by backend, the driver would schedule
>>>>> a workqueue to send gratituous packet through NETDEV_NOTIFY_PEERS.
>>>>>
>>>>> This feature is negotiated through bit VIRTIO_NET_F_GUEST_ANNOUNCE.
>>>>>
>>>>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>>>>
>>>> This seems like a huge layering violation. Imagine this in real
>>>> hardware, for example.
>>>
>>> commits 06c4648d46d1b757d6b9591a86810be79818b60c
>>> and 99606477a5888b0ead0284fecb13417b1da8e3af
>>> document the need for this:
>>>
>>> NETDEV_NOTIFY_PEERS notifier indicates that a device moved to a
>>> different physical link.
>>> and
>>> In real hardware such notifications are only
>>> generated when the device comes up or the address changes.
>>>
>>> So hypervisor could get the same behaviour by sending link up/down
>>> events, this is just an optimization so guest won't do
>>> unecessary stuff like try to reconfigure an IP address.
>>>
>>>
>>> Maybe LOCATION_CHANGE would be a better name?
>>>
>>
>> ANNOUNCE_SELF?
>
> It would be nice to formulate what kind of event
> are we notifying the guest about.
> The announce part of it is really up to the guest, isn't it?
>
Right.
>>>
>>>> There may be a good reason why virtual devices might want this kind of
>>>> reconfiguration cheat, which is unnecessary for normal machines,
>>>
>>> I think yes, the difference with real hardware is guest can change
>>> location without link getting dropped.
>>> FWIW, Xen seems to use this capability too.
>>
>> So does ms netvsc.
>>
>>>
>>>> but
>>>> it'd have to be spelled out clearly in the spec to justify it...
>>>>
>>>> Cheers,
>>>> Rusty.
>>>
>>> Agree, and I'd like to see the spec too. The interface seems
>>> to involve the guest clearing the status bit when it detects
>>> an event?
>>
>> I would describe this in spec. The interface need guest to clear the
>> status bit, this would let the back-end know it has finished the work as
>> we may need to send the gratuitous packets many times.
>>
>>>
>>> Also - how does it interact with the link up event?
>>> We probably don't want to schedule this when we detect
>>> a link status change or during initialization, as
>>> this patch seems to do? What if link goes down
>>> while the work is running? Is that OK?
>>>
>>
>> Looks like there's are duplications if guest enable arp_notify vm is
>> started,
>
> How hard would it be to avoid these duplicates?
Not hard, it could be done in backend by distinguishing the reason :
fresh start or cont after migration or stop.
>
>> but we need to handle the situation that resuming a stopped
>> virtual machine.
>>
>> For the link down race, I don't see any real issue, either dropping or
>> queued.
>
> For example, you do
> unregister_netdev(vi->dev);
> cancel_work_sync(&vi->announce);
>
> which looks scary as announce seems to use the netdev.
>
oops, it's a bug, I would fix it. Thanks
^ permalink raw reply
* Re: [Qemu-devel] [RFC v2 PATCH 4/4] virtio-net: notify guest to annouce itself
From: Stefan Hajnoczi @ 2011-10-26 6:19 UTC (permalink / raw)
To: Jason Wang
Cc: aliguori, quintela, jan.kiszka, mst, qemu-devel, blauwirbel,
pbonzini, rusty, kvm, netdev
In-Reply-To: <20111022053907.21526.81697.stgit@dhcp-8-146.nay.redhat.com>
On Fri, Oct 21, 2011 at 10:39 PM, Jason Wang <jasowang@redhat.com> wrote:
> diff --git a/hw/virtio-net.c b/hw/virtio-net.c
> index 8c2f460..7f844e7 100644
> --- a/hw/virtio-net.c
> +++ b/hw/virtio-net.c
> @@ -95,6 +95,10 @@ static void virtio_net_set_config(VirtIODevice *vdev, const uint8_t *config)
> memcpy(n->mac, netcfg.mac, ETH_ALEN);
> qemu_format_nic_info_str(&n->nic->nc, n->mac);
> }
> +
> + if (memcmp(&netcfg.status, &n->status, sizeof(n->status))) {
> + memcpy(&n->status, &netcfg.status, sizeof(n->status));
> + }
The memcpy() can be done unconditionally.
> }
>
> static bool virtio_net_started(VirtIONet *n, uint8_t status)
> @@ -227,7 +231,7 @@ static uint32_t virtio_net_get_features(VirtIODevice *vdev, uint32_t features)
> {
> VirtIONet *n = to_virtio_net(vdev);
>
> - features |= (1 << VIRTIO_NET_F_MAC);
> + features |= (1 << VIRTIO_NET_F_MAC | 1 << VIRTIO_NET_F_GUEST_ANNOUNCE);
>
> if (peer_has_vnet_hdr(n)) {
> tap_using_vnet_hdr(n->nic->nc.peer, 1);
> @@ -983,6 +987,19 @@ static void virtio_net_cleanup(VLANClientState *nc)
> n->nic = NULL;
> }
>
> +static int virtio_net_announce(VLANClientState *nc)
> +{
> + VirtIONet *n = DO_UPCAST(NICState, nc, nc)->opaque;
> +
> + if (n->vdev.guest_features & (0x1 << VIRTIO_NET_F_GUEST_ANNOUNCE)) {
> + n->status |= VIRITO_NET_S_ANNOUNCE;
Typo, should be VIRTIO_NET_S_ANNOUNCE.
Stefan
^ permalink raw reply
* Re: [PATCH net-next] ipv4: use IS_ENABLED() macro to cleanup code
From: Eric Dumazet @ 2011-10-26 6:34 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20111025.193037.1312584002692004360.davem@davemloft.net>
Le mardi 25 octobre 2011 à 19:30 -0400, David Miller a écrit :
> net/ipv4/ip_sockglue.c: In function ‘do_ip_setsockopt’:
> net/ipv4/ip_sockglue.c:523:29: error: ‘LOOPBACK4_IPV6’ undeclared (first use in this function)
> net/ipv4/ip_sockglue.c:523:29: note: each undeclared identifier is reported only once for each function it appears in
>
> This fails because ip_sockglue.c guards the net/transp_v6.h header
> inclusion with a real CPP guard.
Yep, it seems compiler was not able to perform optimisation of dead
code.
if (IS_ENABLED(CONFIG_IPV6) &&
(sk->sk_family == PF_INET ||
(!((1 << sk->sk_state) &
(TCPF_LISTEN | TCPF_CLOSE)) &&
inet->inet_daddr != LOOPBACK4_IPV6))) {
Never mind, this will wait 3.3 :)
^ permalink raw reply
* [PATCH] ipv6: tcp: fix TCLASS value in ACK messages sent from TIME_WAIT
From: Eric Dumazet @ 2011-10-26 6:42 UTC (permalink / raw)
To: David Miller; +Cc: netdev
commit 66b13d99d96a (ipv4: tcp: fix TOS value in ACK messages sent from
TIME_WAIT) fixed IPv4 only.
This part is for the IPv6 side.
We alias tw_tclass and tw_tos, if socket family is INET6.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
---
include/net/inet_timewait_sock.h | 1 +
include/net/ipv6.h | 3 ++-
net/ipv4/tcp_minisocks.c | 1 +
net/ipv6/inet6_connection_sock.c | 2 +-
net/ipv6/ip6_output.c | 7 ++-----
net/ipv6/tcp_ipv6.c | 17 +++++++++--------
6 files changed, 16 insertions(+), 15 deletions(-)
diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h
index 180231c..f91a1fb 100644
--- a/include/net/inet_timewait_sock.h
+++ b/include/net/inet_timewait_sock.h
@@ -134,6 +134,7 @@ struct inet_timewait_sock {
struct inet_bind_bucket *tw_tb;
struct hlist_node tw_death_node;
};
+#define tw_tclass tw_tos
static inline void inet_twsk_add_node_rcu(struct inet_timewait_sock *tw,
struct hlist_nulls_head *list)
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 3b5ac1f..a366a8a 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -486,7 +486,8 @@ extern int ip6_rcv_finish(struct sk_buff *skb);
extern int ip6_xmit(struct sock *sk,
struct sk_buff *skb,
struct flowi6 *fl6,
- struct ipv6_txoptions *opt);
+ struct ipv6_txoptions *opt,
+ int tclass);
extern int ip6_nd_hdr(struct sock *sk,
struct sk_buff *skb,
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 85a2fbe..66363b6 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -345,6 +345,7 @@ void tcp_time_wait(struct sock *sk, int state, int timeo)
tw6 = inet6_twsk((struct sock *)tw);
ipv6_addr_copy(&tw6->tw_v6_daddr, &np->daddr);
ipv6_addr_copy(&tw6->tw_v6_rcv_saddr, &np->rcv_saddr);
+ tw->tw_tclass = np->tclass;
tw->tw_ipv6only = np->ipv6only;
}
#endif
diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c
index 2916200..fee46d5 100644
--- a/net/ipv6/inet6_connection_sock.c
+++ b/net/ipv6/inet6_connection_sock.c
@@ -248,7 +248,7 @@ int inet6_csk_xmit(struct sk_buff *skb, struct flowi *fl_unused)
/* Restore final destination back after routing done */
ipv6_addr_copy(&fl6.daddr, &np->daddr);
- res = ip6_xmit(sk, skb, &fl6, np->opt);
+ res = ip6_xmit(sk, skb, &fl6, np->opt, np->tclass);
rcu_read_unlock();
return res;
}
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 1c9bf8b..ff30047 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -180,7 +180,7 @@ int ip6_output(struct sk_buff *skb)
*/
int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi6 *fl6,
- struct ipv6_txoptions *opt)
+ struct ipv6_txoptions *opt, int tclass)
{
struct net *net = sock_net(sk);
struct ipv6_pinfo *np = inet6_sk(sk);
@@ -190,7 +190,6 @@ int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi6 *fl6,
u8 proto = fl6->flowi6_proto;
int seg_len = skb->len;
int hlimit = -1;
- int tclass = 0;
u32 mtu;
if (opt) {
@@ -228,10 +227,8 @@ int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi6 *fl6,
/*
* Fill in the IPv6 header
*/
- if (np) {
- tclass = np->tclass;
+ if (np)
hlimit = np->hop_limit;
- }
if (hlimit < 0)
hlimit = ip6_dst_hoplimit(dst);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index c8683fc..10b2b31 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -513,7 +513,7 @@ static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req,
__tcp_v6_send_check(skb, &treq->loc_addr, &treq->rmt_addr);
ipv6_addr_copy(&fl6.daddr, &treq->rmt_addr);
- err = ip6_xmit(sk, skb, &fl6, opt);
+ err = ip6_xmit(sk, skb, &fl6, opt, np->tclass);
err = net_xmit_eval(err);
}
@@ -979,7 +979,7 @@ static int tcp6_gro_complete(struct sk_buff *skb)
}
static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
- u32 ts, struct tcp_md5sig_key *key, int rst)
+ u32 ts, struct tcp_md5sig_key *key, int rst, u8 tclass)
{
const struct tcphdr *th = tcp_hdr(skb);
struct tcphdr *t1;
@@ -1060,7 +1060,7 @@ static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
dst = ip6_dst_lookup_flow(ctl_sk, &fl6, NULL, false);
if (!IS_ERR(dst)) {
skb_dst_set(buff, dst);
- ip6_xmit(ctl_sk, buff, &fl6, NULL);
+ ip6_xmit(ctl_sk, buff, &fl6, NULL, tclass);
TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS);
if (rst)
TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS);
@@ -1093,13 +1093,13 @@ static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
ack_seq = ntohl(th->seq) + th->syn + th->fin + skb->len -
(th->doff << 2);
- tcp_v6_send_response(skb, seq, ack_seq, 0, 0, key, 1);
+ tcp_v6_send_response(skb, seq, ack_seq, 0, 0, key, 1, 0);
}
static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32 ts,
- struct tcp_md5sig_key *key)
+ struct tcp_md5sig_key *key, u8 tclass)
{
- tcp_v6_send_response(skb, seq, ack, win, ts, key, 0);
+ tcp_v6_send_response(skb, seq, ack, win, ts, key, 0, tclass);
}
static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb)
@@ -1109,7 +1109,8 @@ static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb)
tcp_v6_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt,
tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale,
- tcptw->tw_ts_recent, tcp_twsk_md5_key(tcptw));
+ tcptw->tw_ts_recent, tcp_twsk_md5_key(tcptw),
+ tw->tw_tclass);
inet_twsk_put(tw);
}
@@ -1118,7 +1119,7 @@ static void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb,
struct request_sock *req)
{
tcp_v6_send_ack(skb, tcp_rsk(req)->snt_isn + 1, tcp_rsk(req)->rcv_isn + 1, req->rcv_wnd, req->ts_recent,
- tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr));
+ tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr), 0);
}
^ permalink raw reply related
* Re: [PATCH V2 2/4] MIPS: Add board support for Loongson1B
From: Giuseppe CAVALLARO @ 2011-10-26 7:20 UTC (permalink / raw)
To: Kelvin Cheung
Cc: Wu Zhangjin, linux-mips, linux-kernel, ralf, r0bertz, netdev
In-Reply-To: <CAJhJPsU3KMNuGcKaKf3H4pUCeu1+vnEz4DgO_jbm+wDQb1G4OA@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 4725 bytes --]
Hello Kelvin
On 10/26/2011 6:27 AM, Kelvin Cheung wrote:
> Hi Giuseppe,
>
> This patch works well on Loongson1B platform except one thing.
> The rx checksum offload of normal descriptor is disabled by default.
> So, I enabled this functon. And one minor tweak is added to your patch.
> What about your opinion?
Yes, I had not enabled the rx coe. I'm resending your patch (v3) where I
fixed a problem. Old mac10/100 has no rx csum in HW. So I added an extra
check.
Let me consider it the final version. ;-)
Thanks a lot for you effort.
I'll send it for net-next now.
Regards
Peppe
>
> BTW, tx checksum insertion works now.
>
> 2011/10/25, Giuseppe CAVALLARO <peppe.cavallaro@st.com>:
>> On 10/25/2011 9:09 AM, Giuseppe CAVALLARO wrote:
>>> On 10/25/2011 4:12 AM, Kelvin Cheung wrote:
>>>> 2011/10/24, Giuseppe CAVALLARO <peppe.cavallaro@st.com>:
>>>>> On 10/24/2011 4:05 PM, Kelvin Cheung wrote:
>>>>>> 2011/10/24, Giuseppe CAVALLARO <peppe.cavallaro@st.com>:
>>>>>>> Hello Kelvin.
>>>>>>>
>>>>>>> On 10/24/2011 12:36 PM, Kelvin Cheung wrote:
>>>>>>>
>>>>>>> [snip]
>>>>>>>
>>>>>>>> According to datasheet of Loongson 1B, the buffer size in RX/TX
>>>>>>>> descriptor is only 2KB. So the Loongson1B's GMAC could not handle
>>>>>>>> jumbo frames. And the second buffer is useless in this case. Am I
>>>>>>>> right? Is there a better way than ifdef CONFIG_MACH_LOONGSON1 to
>>>>>>>> avoid duplicate code?
>>>>>>>
>>>>>>> Sorry for my misunderstanding.
>>>>>>>
>>>>>>> I think you have to use the normal descriptor and remove the enh_desc
>>>>>>> from the platform w/o modifying the driver at all.
>>>>>>>
>>>>>>> The driver will be able to select/configure all automatically (also
>>>>>>> jumbo).
>>>>>>>
>>>>>>> Let me know.
>>>>>>
>>>>>> That's the problem.
>>>>>> The bitfield definition of Loongson1B is also different from normal
>>>>>> descriptor.
>>>>>
>>>>> The problem is not in the Loongson1B gmac.
>>>>
>>>> I found that the bit checksum_insertion is not existed in normal
>>>> descriptor.
>>>>
>>>>> The normal descriptor fields in the stmmac refer to an old synopsys
>>>>> databook.
>>>>
>>>> Could you send me the new databook of Synopsys GMAC?
>>>>
>>>>> New chips have the same structure you have added; so we should fix this
>>>>> in the driver w/o breaking the compatibility for old chips.
>>>>
>>>> Agree.
>>>>
>>>>> I kindly ask you to confirm if the currently normal descriptor structure
>>>>> (w/o your changes) doesn't work on your platform.
>>>>> Did you test it?
>>>>
>>>> Well, the normal descriptor works on my platform except TX checksum
>>>> offload.
>>>
>>> ok! I suspected that.
>>>
>>>
>>>>>> Moreover, I want to enable the TX checksum offload function which is
>>>>>> not supported in normal descriptor.
>>>>>> Any suggestions?
>>>>>
>>>>> It is supported but you have to pass from the platform: tx_coe = 1.
>>>>
>>>> I noticed that the flag csum_insertion is passed to
>>>> ndesc_prepare_tx_desc() in stmmac_xmit(). But ndesc_prepare_tx_desc()
>>>> just ignores it.
>>>> In other words, the TX checksum offload function is disabled in normal
>>>> descriptor currently.
>>>>
>>>> Should we fix this problem for normal descriptor?
>>>
>>> Yes, we should. If you agree, I'll update the normal descriptor
>>> structure to yours. This is the normal descriptor used in newer GMAC.
>>> Tx csum will be done for normal descriptors in case of these GMAC
>>> devices and not for old MAC10/100. For the MAC10/100 some bits for
>>> normal descriptors are reserved and won't be used at all.
>>>
>>> I'll also verify that the patch doesn't break the back-compatibility
>>> with old MAC10/100. I have the HW where doing the tests.
>>>
>>> After that, I'll prepare the patch for net-next and for your kernel.
>>
>> Hello Kelvin
>>
>> attached the patch tested on my development kernel.
>> It runs fine on old and new mac devices.
>>
>> Can you try it on your side? Hmm, it is likely it won't apply fine on
>> your tree but you know the changes ;-).
>>
>> If ok, I'll rework it for net-next and send it to the mailing list.
>>
>> Thanks
>> Peppe
>>
>>>
>>>>
>>>>> Peppe
>>>>>>
>>>>>>> Note:
>>>>>>> IIRC, there is a bit difference in case of normal descriptors for
>>>>>>> Synopsys databook newer than the 1.91 (I used for testing this mode).
>>>>>>> In any case, I remember that, on some platforms, the normal
>>>>>>> descriptors
>>>>>>> have been used w/o problems also on these new chip generations.
>>>>>>>
>>>>>>> Peppe
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>> --
>>> 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
>>>
>>
>>
>
>
[-- Attachment #2: 0001-stmmac-update-normal-descriptor-structure-v3.patch --]
[-- Type: text/x-patch, Size: 8441 bytes --]
>From 324c8e56e728046590bae14d14e8ff09978b71b4 Mon Sep 17 00:00:00 2001
From: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Date: Tue, 25 Oct 2011 09:46:41 +0200
Subject: [PATCH (sh-2.6.32.y)] stmmac: update normal descriptor structure
(v3)
This patch updates the normal descriptor structure
to work fine on new GMAC Synopsys chips.
Normal descriptors were designed on the old MAC10/100
databook 1.91 where some bits were reserved: for example
the tx checksum insertion and rx checksum offload.
The patch maintains the back-compatibility with old
MAC devices (tested on STx7109 MAC10/100) and adds new
fields that actually new GMAC devices can use.
For example, STx7109 will pass from the platform
tx_coe = 0, enh_desc = 0, has_gmac = 0.
A platform like Loongson1B (Mips) will pass:
tx_coe = 1, enh_desc = 0, has_gmac = 1.
Thanks to Kelvin, he enhanced the normal descriptors for
GMAC on Loongson1B.
Signed-off-by: Kelvin Cheung <keguang.zhang@gmail.com>
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/stmmac/common.h | 8 +++---
drivers/net/stmmac/descs.h | 31 +++++++++++++++-----------
drivers/net/stmmac/norm_desc.c | 40 ++++++++++++++++++++--------------
drivers/net/stmmac/stmmac_ethtool.c | 8 +++---
drivers/net/stmmac/stmmac_main.c | 4 +-
5 files changed, 51 insertions(+), 40 deletions(-)
diff --git a/drivers/net/stmmac/common.h b/drivers/net/stmmac/common.h
index 9100c10..2cc1192 100644
--- a/drivers/net/stmmac/common.h
+++ b/drivers/net/stmmac/common.h
@@ -49,7 +49,7 @@ struct stmmac_extra_stats {
unsigned long tx_underflow ____cacheline_aligned;
unsigned long tx_carrier;
unsigned long tx_losscarrier;
- unsigned long tx_heartbeat;
+ unsigned long vlan_tag;
unsigned long tx_deferred;
unsigned long tx_vlan;
unsigned long tx_jabber;
@@ -58,9 +58,9 @@ struct stmmac_extra_stats {
unsigned long tx_ip_header_error;
/* Receive errors */
unsigned long rx_desc;
- unsigned long rx_partial;
- unsigned long rx_runt;
- unsigned long rx_toolong;
+ unsigned long sa_filter_fail;
+ unsigned long overflow_error;
+ unsigned long ipc_csum_error;
unsigned long rx_collision;
unsigned long rx_crc;
unsigned long rx_length;
diff --git a/drivers/net/stmmac/descs.h b/drivers/net/stmmac/descs.h
index 63a03e2..9820ec8 100644
--- a/drivers/net/stmmac/descs.h
+++ b/drivers/net/stmmac/descs.h
@@ -25,33 +25,34 @@ struct dma_desc {
union {
struct {
/* RDES0 */
- u32 reserved1:1;
+ u32 payload_csum_error:1;
u32 crc_error:1;
u32 dribbling:1;
u32 mii_error:1;
u32 receive_watchdog:1;
u32 frame_type:1;
u32 collision:1;
- u32 frame_too_long:1;
+ u32 ipc_csum_error:1;
u32 last_descriptor:1;
u32 first_descriptor:1;
- u32 multicast_frame:1;
- u32 run_frame:1;
+ u32 vlan_tag:1;
+ u32 overflow_error:1;
u32 length_error:1;
- u32 partial_frame_error:1;
+ u32 sa_filter_fail:1;
u32 descriptor_error:1;
u32 error_summary:1;
u32 frame_length:14;
- u32 filtering_fail:1;
+ u32 da_filter_fail:1;
u32 own:1;
/* RDES1 */
u32 buffer1_size:11;
u32 buffer2_size:11;
- u32 reserved2:2;
+ u32 reserved1:2;
u32 second_address_chained:1;
u32 end_ring:1;
- u32 reserved3:5;
+ u32 reserved2:5;
u32 disable_ic:1;
+
} rx;
struct {
/* RDES0 */
@@ -91,24 +92,28 @@ struct dma_desc {
u32 underflow_error:1;
u32 excessive_deferral:1;
u32 collision_count:4;
- u32 heartbeat_fail:1;
+ u32 vlan_frame:1;
u32 excessive_collisions:1;
u32 late_collision:1;
u32 no_carrier:1;
u32 loss_carrier:1;
- u32 reserved1:3;
+ u32 payload_error:1;
+ u32 frame_flushed:1;
+ u32 jabber_timeout:1;
u32 error_summary:1;
- u32 reserved2:15;
+ u32 ip_header_error:1;
+ u32 time_stamp_status:1;
+ u32 reserved1:13;
u32 own:1;
/* TDES1 */
u32 buffer1_size:11;
u32 buffer2_size:11;
- u32 reserved3:1;
+ u32 time_stamp_enable:1;
u32 disable_padding:1;
u32 second_address_chained:1;
u32 end_ring:1;
u32 crc_disable:1;
- u32 reserved4:2;
+ u32 checksum_insertion:2;
u32 first_segment:1;
u32 last_segment:1;
u32 interrupt:1;
diff --git a/drivers/net/stmmac/norm_desc.c b/drivers/net/stmmac/norm_desc.c
index f7e8ba7..3955ae7 100644
--- a/drivers/net/stmmac/norm_desc.c
+++ b/drivers/net/stmmac/norm_desc.c
@@ -50,11 +50,12 @@ static int ndesc_get_tx_status(void *data, struct stmmac_extra_stats *x,
stats->collisions += p->des01.tx.collision_count;
ret = -1;
}
- if (unlikely(p->des01.tx.heartbeat_fail)) {
- x->tx_heartbeat++;
- stats->tx_heartbeat_errors++;
- ret = -1;
- }
+
+ if (p->des01.etx.vlan_frame) {
+ CHIP_DBG(KERN_INFO "GMAC TX status: VLAN frame\n");
+ x->tx_vlan++;
+ }
+
if (unlikely(p->des01.tx.deferred))
x->tx_deferred++;
@@ -68,12 +69,12 @@ static int ndesc_get_tx_len(struct dma_desc *p)
/* This function verifies if each incoming frame has some errors
* and, if required, updates the multicast statistics.
- * In case of success, it returns csum_none because the device
- * is not able to compute the csum in HW. */
+ * In case of success, it returns good_frame because the GMAC device
+ * is supposed to be able to compute the csum in HW. */
static int ndesc_get_rx_status(void *data, struct stmmac_extra_stats *x,
struct dma_desc *p)
{
- int ret = csum_none;
+ int ret = good_frame;
struct net_device_stats *stats = (struct net_device_stats *)data;
if (unlikely(p->des01.rx.last_descriptor == 0)) {
@@ -86,12 +87,12 @@ static int ndesc_get_rx_status(void *data, struct stmmac_extra_stats *x,
if (unlikely(p->des01.rx.error_summary)) {
if (unlikely(p->des01.rx.descriptor_error))
x->rx_desc++;
- if (unlikely(p->des01.rx.partial_frame_error))
- x->rx_partial++;
- if (unlikely(p->des01.rx.run_frame))
- x->rx_runt++;
- if (unlikely(p->des01.rx.frame_too_long))
- x->rx_toolong++;
+ if (unlikely(p->des01.rx.sa_filter_fail))
+ x->sa_filter_fail++;
+ if (unlikely(p->des01.rx.overflow_error))
+ x->overflow_error++;
+ if (unlikely(p->des01.rx.ipc_csum_error))
+ x->ipc_csum_error++;
if (unlikely(p->des01.rx.collision)) {
x->rx_collision++;
stats->collisions++;
@@ -113,10 +114,12 @@ static int ndesc_get_rx_status(void *data, struct stmmac_extra_stats *x,
x->rx_mii++;
ret = discard_frame;
}
- if (p->des01.rx.multicast_frame) {
- x->rx_multicast++;
- stats->multicast++;
+#ifdef STMMAC_VLAN_TAG_USED
+ if (p->des01.rx.vlan_tag) {
+ x->vlan_tag++;
+ stats->vlan_tag++;
}
+#endif
return ret;
}
@@ -184,6 +187,9 @@ static void ndesc_prepare_tx_desc(struct dma_desc *p, int is_fs, int len,
{
p->des01.tx.first_segment = is_fs;
norm_set_tx_desc_len(p, len);
+
+ if (likely(csum_flag))
+ p->des01.tx.checksum_insertion = cic_full;
}
static void ndesc_clear_tx_ic(struct dma_desc *p)
diff --git a/drivers/net/stmmac/stmmac_ethtool.c b/drivers/net/stmmac/stmmac_ethtool.c
index d81cf20..6831cf2 100644
--- a/drivers/net/stmmac/stmmac_ethtool.c
+++ b/drivers/net/stmmac/stmmac_ethtool.c
@@ -48,7 +48,7 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(tx_underflow),
STMMAC_STAT(tx_carrier),
STMMAC_STAT(tx_losscarrier),
- STMMAC_STAT(tx_heartbeat),
+ STMMAC_STAT(vlan_tag),
STMMAC_STAT(tx_deferred),
STMMAC_STAT(tx_vlan),
STMMAC_STAT(rx_vlan),
@@ -57,9 +57,9 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(tx_payload_error),
STMMAC_STAT(tx_ip_header_error),
STMMAC_STAT(rx_desc),
- STMMAC_STAT(rx_partial),
- STMMAC_STAT(rx_runt),
- STMMAC_STAT(rx_toolong),
+ STMMAC_STAT(sa_filter_fail),
+ STMMAC_STAT(overflow_error),
+ STMMAC_STAT(ipc_csum_error),
STMMAC_STAT(rx_collision),
STMMAC_STAT(rx_crc),
STMMAC_STAT(rx_length),
diff --git a/drivers/net/stmmac/stmmac_main.c b/drivers/net/stmmac/stmmac_main.c
index 3a3efb1..18e614b 100644
--- a/drivers/net/stmmac/stmmac_main.c
+++ b/drivers/net/stmmac/stmmac_main.c
@@ -1318,8 +1318,8 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit)
#endif
skb->protocol = eth_type_trans(skb, priv->dev);
- if (unlikely(status == csum_none)) {
- /* always for the old mac 10/100 */
+ if (unlikely(!priv->plat->has_gmac)) {
+ /* No csum for the old mac 10/100 devices */
skb->ip_summed = CHECKSUM_NONE;
netif_receive_skb(skb);
} else {
--
1.7.4.4
^ permalink raw reply related
* [net 0/7] bnx2x: driver and firmware fixes
From: Yaniv Rosner @ 2011-10-26 9:40 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Yaniv Rosner
Hi Dave,
The following patch series describe few link fixes and firmware update.
Please consider applying it to net.
Thanks,
Yaniv
^ permalink raw reply
* [net 2/7] bnx2x: Add link retry to 578xx-KR
From: Yaniv Rosner @ 2011-10-26 9:40 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Yaniv Rosner, Eilon Greenstein
In-Reply-To: <1319622039-10553-1-git-send-email-yanivr@broadcom.com>
This fix solves a problem of no link on 578xx-KR by retrying to link up to
four timer using the periodic function.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 117 ++++++++++++++++++++++++++++++++++------
drivers/net/bnx2x/bnx2x_link.h | 3 +
2 files changed, 104 insertions(+), 16 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index 5777dcd..974a50d 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -261,6 +261,7 @@
#define MAX_PACKET_SIZE (9700)
#define WC_UC_TIMEOUT 100
+#define MAX_KR_LINK_RETRY 4
/**********************************************************/
/* INTERFACE */
@@ -3578,6 +3579,11 @@ static void bnx2x_warpcore_enable_AN_KR(struct bnx2x_phy *phy,
u16 val16 = 0, lane, bam37 = 0;
struct bnx2x *bp = params->bp;
DP(NETIF_MSG_LINK, "Enable Auto Negotiation for KR\n");
+
+ /* Disable Autoneg: re-enable it after adv is done. */
+ bnx2x_cl45_write(bp, phy, MDIO_AN_DEVAD,
+ MDIO_WC_REG_IEEE0BLK_MIICNTL, 0);
+
/* Check adding advertisement for 1G KX */
if (((vars->line_speed == SPEED_AUTO_NEG) &&
(phy->speed_cap_mask & PORT_HW_CFG_SPEED_CAPABILITY_D0_1G)) ||
@@ -3619,9 +3625,6 @@ static void bnx2x_warpcore_enable_AN_KR(struct bnx2x_phy *phy,
bnx2x_cl45_write(bp, phy, MDIO_WC_DEVAD,
MDIO_WC_REG_CL72_USERB0_CL72_2P5_DEF_CTRL,
0x03f0);
- bnx2x_cl45_write(bp, phy, MDIO_WC_DEVAD,
- MDIO_WC_REG_CL72_USERB0_CL72_MISC1_CONTROL,
- 0x383f);
/* Advertised speeds */
bnx2x_cl45_write(bp, phy, MDIO_AN_DEVAD,
@@ -3648,19 +3651,22 @@ static void bnx2x_warpcore_enable_AN_KR(struct bnx2x_phy *phy,
/* Advertise pause */
bnx2x_ext_phy_set_pause(params, phy, vars);
- /* Enable Autoneg */
- bnx2x_cl45_write(bp, phy, MDIO_AN_DEVAD,
- MDIO_WC_REG_IEEE0BLK_MIICNTL, 0x1000);
-
- /* Over 1G - AN local device user page 1 */
- bnx2x_cl45_write(bp, phy, MDIO_WC_DEVAD,
- MDIO_WC_REG_DIGITAL3_UP1, 0x1f);
+ vars->rx_tx_asic_rst = MAX_KR_LINK_RETRY;
bnx2x_cl45_read(bp, phy, MDIO_WC_DEVAD,
MDIO_WC_REG_DIGITAL5_MISC7, &val16);
bnx2x_cl45_write(bp, phy, MDIO_WC_DEVAD,
MDIO_WC_REG_DIGITAL5_MISC7, val16 | 0x100);
+
+ /* Over 1G - AN local device user page 1 */
+ bnx2x_cl45_write(bp, phy, MDIO_WC_DEVAD,
+ MDIO_WC_REG_DIGITAL3_UP1, 0x1f);
+
+ /* Enable Autoneg */
+ bnx2x_cl45_write(bp, phy, MDIO_AN_DEVAD,
+ MDIO_WC_REG_IEEE0BLK_MIICNTL, 0x1000);
+
}
static void bnx2x_warpcore_set_10G_KR(struct bnx2x_phy *phy,
@@ -4129,6 +4135,85 @@ static int bnx2x_is_sfp_module_plugged(struct bnx2x_phy *phy,
else
return 0;
}
+static int bnx2x_warpcore_get_sigdet(struct bnx2x_phy *phy,
+ struct link_params *params)
+{
+ u16 gp2_status_reg0, lane;
+ struct bnx2x *bp = params->bp;
+
+ lane = bnx2x_get_warpcore_lane(phy, params);
+
+ bnx2x_cl45_read(bp, phy, MDIO_WC_DEVAD, MDIO_WC_REG_GP2_STATUS_GP_2_0,
+ &gp2_status_reg0);
+
+ return (gp2_status_reg0 >> (8+lane)) & 0x1;
+}
+
+static void bnx2x_warpcore_config_runtime(struct bnx2x_phy *phy,
+ struct link_params *params,
+ struct link_vars *vars)
+{
+ struct bnx2x *bp = params->bp;
+ u32 serdes_net_if;
+ u16 gp_status1 = 0, lnkup = 0, lnkup_kr = 0;
+ u16 lane = bnx2x_get_warpcore_lane(phy, params);
+
+ vars->turn_to_run_wc_rt = vars->turn_to_run_wc_rt ? 0 : 1;
+
+ if (!vars->turn_to_run_wc_rt)
+ return;
+
+ /* return if there is no link partner */
+ if (!(bnx2x_warpcore_get_sigdet(phy, params))) {
+ DP(NETIF_MSG_LINK, "bnx2x_warpcore_get_sigdet false\n");
+ return;
+ }
+
+ if (vars->rx_tx_asic_rst) {
+ serdes_net_if = (REG_RD(bp, params->shmem_base +
+ offsetof(struct shmem_region, dev_info.
+ port_hw_config[params->port].default_cfg)) &
+ PORT_HW_CFG_NET_SERDES_IF_MASK);
+
+ switch (serdes_net_if) {
+ case PORT_HW_CFG_NET_SERDES_IF_KR:
+ /* Do we get link yet? */
+ bnx2x_cl45_read(bp, phy, MDIO_WC_DEVAD, 0x81d1,
+ &gp_status1);
+ lnkup = (gp_status1 >> (8+lane)) & 0x1;/* 1G */
+ /*10G KR*/
+ lnkup_kr = (gp_status1 >> (12+lane)) & 0x1;
+
+ DP(NETIF_MSG_LINK,
+ "gp_status1 0x%x\n", gp_status1);
+
+ if (lnkup_kr || lnkup) {
+ vars->rx_tx_asic_rst = 0;
+ DP(NETIF_MSG_LINK,
+ "link up, rx_tx_asic_rst 0x%x\n",
+ vars->rx_tx_asic_rst);
+ } else {
+ /*reset the lane to see if link comes up.*/
+ bnx2x_warpcore_reset_lane(bp, phy, 1);
+ bnx2x_warpcore_reset_lane(bp, phy, 0);
+
+ /* restart Autoneg */
+ bnx2x_cl45_write(bp, phy, MDIO_AN_DEVAD,
+ MDIO_WC_REG_IEEE0BLK_MIICNTL, 0x1200);
+
+ vars->rx_tx_asic_rst--;
+ DP(NETIF_MSG_LINK, "0x%x retry left\n",
+ vars->rx_tx_asic_rst);
+ }
+ break;
+
+ default:
+ break;
+ }
+
+ } /*params->rx_tx_asic_rst*/
+
+}
static void bnx2x_warpcore_config_init(struct bnx2x_phy *phy,
struct link_params *params,
@@ -12331,11 +12416,6 @@ void bnx2x_period_func(struct link_params *params, struct link_vars *vars)
{
struct bnx2x *bp = params->bp;
u16 phy_idx;
- if (!params) {
- DP(NETIF_MSG_LINK, "Uninitialized params !\n");
- return;
- }
-
for (phy_idx = INT_PHY; phy_idx < MAX_PHYS; phy_idx++) {
if (params->phy[phy_idx].flags & FLAGS_TX_ERROR_CHECK) {
bnx2x_set_aer_mmd(params, ¶ms->phy[phy_idx]);
@@ -12344,8 +12424,13 @@ void bnx2x_period_func(struct link_params *params, struct link_vars *vars)
}
}
- if (CHIP_IS_E3(bp))
+ if (CHIP_IS_E3(bp)) {
+ struct bnx2x_phy *phy = ¶ms->phy[INT_PHY];
+ bnx2x_set_aer_mmd(params, phy);
bnx2x_check_over_curr(params, vars);
+ bnx2x_warpcore_config_runtime(phy, params, vars);
+ }
+
}
u8 bnx2x_hw_lock_required(struct bnx2x *bp, u32 shmem_base, u32 shmem2_base)
diff --git a/drivers/net/bnx2x/bnx2x_link.h b/drivers/net/bnx2x/bnx2x_link.h
index c12db6d..2a46e63 100644
--- a/drivers/net/bnx2x/bnx2x_link.h
+++ b/drivers/net/bnx2x/bnx2x_link.h
@@ -303,6 +303,9 @@ struct link_vars {
#define PERIODIC_FLAGS_LINK_EVENT 0x0001
u32 aeu_int_mask;
+ u8 rx_tx_asic_rst;
+ u8 turn_to_run_wc_rt;
+ u16 rsrv2;
};
/***********************************************************/
--
1.7.7.1
^ permalink raw reply related
* [net 3/7] bnx2x: Fix RX/TX problem caused by the MAC layer
From: Yaniv Rosner @ 2011-10-26 9:40 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Yaniv Rosner, Eilon Greenstein
In-Reply-To: <1319622039-10553-1-git-send-email-yanivr@broadcom.com>
This patch fixes a problem in which the host stops receiving data after
restarting the interface. This issue is caused by combination of incorrect
data path tap closure, along with missing MAC reset.
Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
---
drivers/net/bnx2x/bnx2x_link.c | 49 ++++++++++++++++++++++++++++-----------
1 files changed, 35 insertions(+), 14 deletions(-)
diff --git a/drivers/net/bnx2x/bnx2x_link.c b/drivers/net/bnx2x/bnx2x_link.c
index 974a50d..ea2a11d 100644
--- a/drivers/net/bnx2x/bnx2x_link.c
+++ b/drivers/net/bnx2x/bnx2x_link.c
@@ -1494,6 +1494,18 @@ static void bnx2x_set_xumac_nig(struct link_params *params,
NIG_REG_P0_MAC_PAUSE_OUT_EN, tx_pause_en);
}
+static void bnx2x_umac_disable(struct link_params *params)
+{
+ u32 umac_base = params->port ? GRCBASE_UMAC1 : GRCBASE_UMAC0;
+ struct bnx2x *bp = params->bp;
+ if (!(REG_RD(bp, MISC_REG_RESET_REG_2) &
+ (MISC_REGISTERS_RESET_REG_2_UMAC0 << params->port)))
+ return;
+
+ /* Disable RX and TX */
+ REG_WR(bp, umac_base + UMAC_REG_COMMAND_CONFIG, 0);
+}
+
static void bnx2x_umac_enable(struct link_params *params,
struct link_vars *vars, u8 lb)
{
@@ -1603,8 +1615,9 @@ static u8 bnx2x_is_4_port_mode(struct bnx2x *bp)
}
/* Define the XMAC mode */
-static void bnx2x_xmac_init(struct bnx2x *bp, u32 max_speed)
+static void bnx2x_xmac_init(struct link_params *params, u32 max_speed)
{
+ struct bnx2x *bp = params->bp;
u32 is_port4mode = bnx2x_is_4_port_mode(bp);
/**
@@ -1614,10 +1627,11 @@ static void bnx2x_xmac_init(struct bnx2x *bp, u32 max_speed)
* ports of the path
**/
- if (is_port4mode && (REG_RD(bp, MISC_REG_RESET_REG_2) &
+ if ((CHIP_NUM(bp) == CHIP_NUM_57840) &&
+ (REG_RD(bp, MISC_REG_RESET_REG_2) &
MISC_REGISTERS_RESET_REG_2_XMAC)) {
- DP(NETIF_MSG_LINK, "XMAC already out of reset"
- " in 4-port mode\n");
+ DP(NETIF_MSG_LINK,
+ "XMAC already out of reset in 4-port mode\n");
return;
}
@@ -1681,10 +1695,6 @@ static void bnx2x_xmac_disable(struct link_params *params)
(pfc_ctrl | (1<<1)));
DP(NETIF_MSG_LINK, "Disable XMAC on port %x\n", port);
REG_WR(bp, xmac_base + XMAC_REG_CTRL, 0);
- usleep_range(1000, 1000);
- bnx2x_set_xumac_nig(params, 0, 0);
- REG_WR(bp, xmac_base + XMAC_REG_CTRL,
- XMAC_CTRL_REG_SOFT_RESET);
}
}
@@ -1697,7 +1707,7 @@ static int bnx2x_xmac_enable(struct link_params *params,
xmac_base = (params->port) ? GRCBASE_XMAC1 : GRCBASE_XMAC0;
- bnx2x_xmac_init(bp, vars->line_speed);
+ bnx2x_xmac_init(params, vars->line_speed);
/*
* This register determines on which events the MAC will assert
@@ -6309,8 +6319,10 @@ static int bnx2x_update_link_down(struct link_params *params,
MISC_REGISTERS_RESET_REG_2_CLEAR,
(MISC_REGISTERS_RESET_REG_2_RST_BMAC0 << port));
}
- if (CHIP_IS_E3(bp))
+ if (CHIP_IS_E3(bp)) {
bnx2x_xmac_disable(params);
+ bnx2x_umac_disable(params);
+ }
return 0;
}
@@ -11802,8 +11814,10 @@ int bnx2x_link_reset(struct link_params *params, struct link_vars *vars,
/* Stop BigMac rx */
if (!CHIP_IS_E3(bp))
bnx2x_bmac_rx_disable(bp, port);
- else
+ else {
bnx2x_xmac_disable(params);
+ bnx2x_umac_disable(params);
+ }
/* disable emac */
if (!CHIP_IS_E3(bp))
REG_WR(bp, NIG_REG_NIG_EMAC0_EN + port*4, 0);
@@ -11841,14 +11855,21 @@ int bnx2x_link_reset(struct link_params *params, struct link_vars *vars,
if (params->phy[INT_PHY].link_reset)
params->phy[INT_PHY].link_reset(
¶ms->phy[INT_PHY], params);
- /* reset BigMac */
- REG_WR(bp, GRCBASE_MISC + MISC_REGISTERS_RESET_REG_2_CLEAR,
- (MISC_REGISTERS_RESET_REG_2_RST_BMAC0 << port));
/* disable nig ingress interface */
if (!CHIP_IS_E3(bp)) {
+ /* reset BigMac */
+ REG_WR(bp, GRCBASE_MISC + MISC_REGISTERS_RESET_REG_2_CLEAR,
+ (MISC_REGISTERS_RESET_REG_2_RST_BMAC0 << port));
REG_WR(bp, NIG_REG_BMAC0_IN_EN + port*4, 0);
REG_WR(bp, NIG_REG_EMAC0_IN_EN + port*4, 0);
+ } else {
+ u32 xmac_base = (params->port) ? GRCBASE_XMAC1 : GRCBASE_XMAC0;
+ bnx2x_set_xumac_nig(params, 0, 0);
+ if (REG_RD(bp, MISC_REG_RESET_REG_2) &
+ MISC_REGISTERS_RESET_REG_2_XMAC)
+ REG_WR(bp, xmac_base + XMAC_REG_CTRL,
+ XMAC_CTRL_REG_SOFT_RESET);
}
vars->link_up = 0;
vars->phy_flags = 0;
--
1.7.7.1
^ 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