- 16 Jan, 2014 26 commits
-
-
Srinivas Kandagatla authored
This patch adds support to "max-speed" property which is a standard Ethernet device tree property. max-speed specifies maximum speed (specified in megabits per second) supported the device. Depending on the clocking schemes some of the boards can only support few link speeds, so having a way to limit the link speed in the mac driver would allow such setups to work reliably. Without this patch there is no way to tell the driver to limit the link speed. Signed-off-by:
Srinivas Kandagatla <srinivas.kandagatla@st.com> Acked-by:
Giuseppe Cavallaro <peppe.cavallaro@st.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Willy Tarreau says: ==================== Assorted mvneta fixes and improvements this series provides some fixes for a number of issues met with the mvneta driver, then adds some improvements. Patches 1-5 are fixes and would be needed in 3.13 and likely -stable. The next ones are performance improvements and cleanups : - driver lockup when reading stats while sending traffic from multiple CPUs : this obviously only happens on SMP and is the result of missing locking on the driver. The problem was present since the introduction of the driver in 3.8. The first patch performs some changes that are needed for the second one which actually fixes the issue by using per-cpu counters. It could make sense to backport this to the relevant stable versions. - mvneta_tx_timeout calls various functions to reset the NIC, and these functions sleep, which is not allowed here, resulting in a panic. Better completely disable this Tx timeout handler for now since it is never called. The problem was encountered while developing some new features, it's uncertain whether it's possible to reproduce it with regular usage, so maybe a backport to stable is not needed. - replace the Tx timer with a real Tx IRQ. As first reported by Arnaud Ebalard and explained by Eric Dumazet, there is no way this driver can work correctly if it uses a driver to recycle the Tx descriptors. If too many packets are sent at once, the driver quickly ends up with no descriptors (which happens twice as easily in GSO) and has to wait 10ms for recycling its descriptors and being able to send again. Eric has worked around this in the core GSO code. But still when routing traffic or sending UDP packets, the limitation is very visible. Using Tx IRQs allows Tx descriptors to be recycled when sent. The coalesce value is still configurable using ethtool. This fix turns the UDP send bitrate from 134 Mbps to 987 Mbps (ie: line rate). It's made of two patches, one to add the relevant bits from the original Marvell's driver, and another one to implement the change. I don't know if it should be backported to stable, as the bug only causes poor performance. - Patches 6..8 are essentially cleanups, code deduplication and minor optimizations for not re-fetching a value we already have (status). - patch 9 changes the prefetch of Rx descriptor from current one to next one. In benchmarks, it results in about 1% general performance increase on HTTP traffic, probably because prefetching the current descriptor does not leave enough time between the start of prefetch and its usage. - patch 10 implements support for build_skb() on Rx path. The driver now preallocates frags instead of skbs and builds an skb just before delivering it. This results in a 2% performance increase on HTTP traffic, and up to 5% on small packet Rx rate. - patch 11 implements rx_copybreak for small packets (256 bytes). It avoids a dma_map_single()/dma_unmap_single() and increases the Rx rate by 16.4%, from 486kpps to 573kpps. Further improvements up to 711kpps are possible depending how the DMA is used. - patches 12 and 13 are extra cleanups made possible by some of the simplifications above. ==================== Signed-off-by: -
Arnaud Ebalard authored
The function return parameter is not used in mvneta_tx_done_gbe(), where the function is called. This patch makes the function return void. Reviewed-by:
Willy Tarreau <w@1wt.eu> Signed-off-by:
Arnaud Ebalard <arno@natisbad.org> Signed-off-by:
-
Arnaud Ebalard authored
mvneta_tx_done_gbe() return value and third parameter are no more used. This patch changes the function prototype and removes a useless variable where the function is called. Reviewed-by:
-
willy tarreau authored
calling dma_map_single()/dma_unmap_single() is quite expensive compared to copying a small packet. So let's copy short frames and keep the buffers mapped. We set the limit to 256 bytes which seems to give good results both on the XP-GP board and on the AX3/4. The Rx small packet rate increased by 16.4% doing this, from 486kpps to 573kpps. It is worth noting that even the call to the function dma_sync_single_range_for_cpu() is expensive (300 ns) although less than dma_unmap_single(). Without it, the packet rate raises to 711kpps (+24% more). Thus on systems where coherency from device to CPU is guaranteed by a snoop control unit, this patch should provide even more gains, and probably rx_copybreak could be increased. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by:
-
willy tarreau authored
Make use of build_skb() to allocate frags on the RX path. When frag size is lower than a page size, we can use netdev_alloc_frag(), and we fall back to kmalloc() for larger sizes. The frag size is stored into the mvneta_port struct. The alloc/free functions check the frag size to decide what alloc/ free method to use. MTU changes are safe because the MTU change function stops the device and clears the queues before applying the change. With this patch, I observed a reproducible 2% performance improvement on HTTP-based benchmarks, and 5% on small packet RX rate. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by:
-
willy tarreau authored
Currently, the mvneta driver tries to prefetch the current Rx descriptor during read. Tests have shown that prefetching the next one instead increases general performance by about 1% on HTTP traffic. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by:
-
willy tarreau authored
At several places, we already know the value of the rx status but we call functions which dereference the pointer again to get it and don't need the descriptor for anything else. Simplify this task by replacing the rx desc pointer by the status word itself. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by:
-
willy tarreau authored
Make mvneta_rxq_fill() use mvneta_rx_refill() instead of using duplicate code. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by:
-
willy tarreau authored
Currently, mvneta_txq_bufs_free() calls mvneta_tx_done_policy() with a non-null cause to retrieve the pointer to the next queue to process. There are useless tests on the return queue number and on the pointer, all of which are well defined within a known limited set. This code path is fast, although not critical. Removing 3 tests here that the compiler could not optimize (verified) is always desirable. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by:
-
willy tarreau authored
Right now the mvneta driver doesn't handle Tx IRQ, and relies on two mechanisms to flush Tx descriptors : a flush at the end of mvneta_tx() and a timer. If a burst of packets is emitted faster than the device can send them, then the queue is stopped until next wake-up of the timer 10ms later. This causes jerky output traffic with bursts and pauses, making it difficult to reach line rate with very few streams. A test on UDP traffic shows that it's not possible to go beyond 134 Mbps / 12 kpps of outgoing traffic with 1500-bytes IP packets. Routed traffic tends to observe pauses as well if the traffic is bursty, making it even burstier after the wake-up. It seems that this feature was inherited from the original driver but nothing there mentions any reason for not using the interrupt instead, which the chip supports. Thus, this patch enables Tx interrupts and removes the timer. It does the two at once because it's not really possible to make the two mechanisms coexist, so a split patch doesn't make sense. First tests performed on a Mirabox (Armada 370) show that less CPU seems to be used when sending traffic. One reason might be that we now call the mvneta_tx_done_gbe() with a mask indicating which queues have been done instead of looping over all of them. The same UDP test above now happily reaches 987 Mbps / 87.7 kpps. Single-stream TCP traffic can now more easily reach line rate. HTTP transfers of 1 MB objects over a single connection went from 730 to 840 Mbps. It is even possible to go significantly higher (>900 Mbps) by tweaking tcp_tso_win_divisor. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Cc: Arnaud Ebalard <arno@natisbad.org> Cc: Eric Dumazet <eric.dumazet@gmail.com> Tested-by:
-
willy tarreau authored
Marvell has not published the chip's datasheet yet, so it's very hard to find the relevant bits to manipulate to change the IRQ behaviour. Fortunately, these bits are described in the proprietary LSP patch set which is publicly available here : http://www.plugcomputer.org/downloads/mirabox/ So let's put them back in the driver in order to reduce the burden of current and future maintenance. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Tested-by: -
willy tarreau authored
If a queue timeout is reported, we can oops because of some schedules while the caller is atomic, as shown below : mvneta d0070000.ethernet eth0: tx timeout BUG: scheduling while atomic: bash/1528/0x00000100 Modules linked in: slhttp_ethdiv(C) [last unloaded: slhttp_ethdiv] CPU: 2 PID: 1528 Comm: bash Tainted: G WC 3.13.0-rc4-mvebu-nf #180 [<c0011bd9>] (unwind_backtrace+0x1/0x98) from [<c000f1ab>] (show_stack+0xb/0xc) [<c000f1ab>] (show_stack+0xb/0xc) from [<c02ad323>] (dump_stack+0x4f/0x64) [<c02ad323>] (dump_stack+0x4f/0x64) from [<c02abe67>] (__schedule_bug+0x37/0x4c) [<c02abe67>] (__schedule_bug+0x37/0x4c) from [<c02ae261>] (__schedule+0x325/0x3ec) [<c02ae261>] (__schedule+0x325/0x3ec) from [<c02adb97>] (schedule_timeout+0xb7/0x118) [<c02adb97>] (schedule_timeout+0xb7/0x118) from [<c0020a67>] (msleep+0xf/0x14) [<c0020a67>] (msleep+0xf/0x14) from [<c01dcbe5>] (mvneta_stop_dev+0x21/0x194) [<c01dcbe5>] (mvneta_stop_dev+0x21/0x194) from [<c01dcfe9>] (mvneta_tx_timeout+0x19/0x24) [<c01dcfe9>] (mvneta_tx_timeout+0x19/0x24) from [<c024afc7>] (dev_watchdog+0x18b/0x1c4) [<c024afc7>] (dev_watchdog+0x18b/0x1c4) from [<c0020b53>] (call_timer_fn.isra.27+0x17/0x5c) [<c0020b53>] (call_timer_fn.isra.27+0x17/0x5c) from [<c0020cad>] (run_timer_softirq+0x115/0x170) [<c0020cad>] (run_timer_softirq+0x115/0x170) from [<c001ccb9>] (__do_softirq+0xbd/0x1a8) [<c001ccb9>] (__do_softirq+0xbd/0x1a8) from [<c001cfad>] (irq_exit+0x61/0x98) [<c001cfad>] (irq_exit+0x61/0x98) from [<c000d4bf>] (handle_IRQ+0x27/0x60) [<c000d4bf>] (handle_IRQ+0x27/0x60) from [<c000843b>] (armada_370_xp_handle_irq+0x33/0xc8) [<c000843b>] (armada_370_xp_handle_irq+0x33/0xc8) from [<c000fba9>] (__irq_usr+0x49/0x60) Ben Hutchings attempted to propose a better fix consisting in using a scheduled work for this, but while it fixed this panic, it caused other random freezes and panics proving that the reset sequence in the driver is unreliable and that additional fixes should be investigated. When sending multiple streams over a link limited to 100 Mbps, Tx timeouts happen from time to time, and the driver correctly recovers only when the function is disabled. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Cc: Ben Hutchings <ben@decadent.org.uk> Tested-by:
-
willy tarreau authored
Stats writers are mvneta_rx() and mvneta_tx(). They don't lock anything when they update the stats, and as a result, it randomly happens that the stats freeze on SMP if two updates happen during stats retrieval. This is very easily reproducible by starting two HTTP servers and binding each of them to a different CPU, then consulting /proc/net/dev in loops during transfers, the interface should immediately lock up. This issue also randomly happens upon link state changes during transfers, because the stats are collected in this situation, but it takes more attempts to reproduce it. The comments in netdevice.h suggest using per_cpu stats instead to get rid of this issue. This patch implements this. It merges both rx_stats and tx_stats into a single "stats" member with a single syncp. Both mvneta_rx() and mvneta_rx() now only update the a single CPU's counters. In turn, mvneta_get_stats64() does the summing by iterating over all CPUs to get their respective stats. With this change, stats are still correct and no more lockup is encountered. Note that this bug was present since the first import of the mvneta driver. It might make sense to backport it to some stable trees. If so, it depends on "d33dc73 net: mvneta: increase the 64-bit rx/tx stats out of the hot path". Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Reviewed-by:
Eric Dumazet <edumazet@google.com> Tested-by:
-
willy tarreau authored
Better count packets and bytes in the stack and on 32 bit then accumulate them at the end for once. This saves two memory writes and two memory barriers per packet. The incoming packet rate was increased by 4.7% on the Openblocks AX3 thanks to this. Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Gregory CLEMENT <gregory.clement@free-electrons.com> Reviewed-by:
-
Paul Gortmaker authored
None of these files are actually using any __init type directives and hence don't need to include <linux/init.h>. Most are just a left over from __devinit and __cpuinit removal, or simply due to code getting copied from one driver to the next. This covers everything under drivers/net except for wireless, which has been submitted separately. Signed-off-by:
Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nftablesDavid S. Miller authored
Pablo Neira Ayuso says: ==================== This small batch contains several Netfilter fixes for your net-next tree, more specifically: * Fix compilation warning in nft_ct in NF_CONNTRACK_MARK is not set, from Kristian Evensen. * Add dependency to IPV6 for NF_TABLES_INET. This one has been reported by the several robots that are testing .config combinations, from Paul Gortmaker. * Fix default base chain policy setting in nf_tables, from myself. ==================== Signed-off-by: -
Jiri Pirko authored
When ndo_neigh_setup is called, the bitfield used by NEIGH_VAR_SET is not initialized yet. This might cause confusion for the people who use NEIGH_VAR_SET in ndo_neigh_setup. So rather introduce NEIGH_VAR_INIT for usage in ndo_neigh_setup. Signed-off-by:
Jiri Pirko <jiri@resnulli.us> Signed-off-by:
-
David S. Miller authored
Aaron Brown says: ==================== Intel Wired LAN Driver Updates This series contains several updates from Alex to ixgbe. To avoid head of line blocking in the event a VF stops cleaning Rx descriptors he makes sure QDE bits are set for a VF before the Rx queues are enabled. To avoid a situation where the head write-back registers can remain set ofter the driver is unloaded he clears them on a VF reset. Alexander Duyck (2): ixgbe: Force QDE via PFQDE for VFs during reset ixgbe: Clear head write-back registers on VF reset ==================== Signed-off-by: -
Alexander Duyck authored
The Tx head write-back registers are not cleared during an FLR or VF reset. As a result a configuration that had head write-back enabled can leave the registers set after the driver is unloaded. If the next driver loaded doesn't use the write-back registers this can lead to a bad configuration where head write-back is enabled, but the driver didn't request it. To avoid this situation the PF should be resetting the Tx head write-back registers when the VF requests a reset. Signed-off-by:
Alexander Duyck <alexander.h.duyck@intel.com> Tested-by:
Phil Schmitt <phillip.j.schmitt@intel.com> Signed-off-by:
Aaron Brown <aaron.f.brown@intel.com> Signed-off-by:
-
Alexander Duyck authored
This change makes it so that the QDE bits are set for a VF before the Rx queues are enabled. As such we avoid head of line blocking in the event that the VF stops cleaning Rx descriptors for whatever reason. Signed-off-by:
-
David S. Miller authored
Thomas Haller says: ==================== ipv6 addrconf: add IFA_F_NOPREFIXROUTE flag to suppress creation of IP6 routes v1 -> v2: add a second commit, handling NOPREFIXROUTE in ip6_del_addr. v2 -> v3: reword commit messages, code comments and some refactoring. v3 -> v4: refactor, rename variables, add enum v4 -> v5: rebase, so that patch applies cleanly to current net-next/master ==================== Signed-off-by: -
Thomas Haller authored
Refactor the deletion/update of prefix routes when removing an address. Now also consider IFA_F_NOPREFIXROUTE and if there is an address present with this flag, to not cleanup the route. Instead, assume that userspace is taking care of this route. Also perform the same cleanup, when userspace changes an existing address to add NOPREFIXROUTE (to an address that didn't have this flag). This is done because when the address was added, a prefix route was created for it. Since the user now wants to handle this route by himself, we cleanup this route. This cleanup of the route is not totally robust. There is no guarantee, that the route we are about to delete was really the one added by the kernel. This behavior does not change by the patch, and in practice it should work just fine. Signed-off-by:
Thomas Haller <thaller@redhat.com> Signed-off-by:
-
Thomas Haller authored
When adding/modifying an IPv6 address, the userspace application needs a way to suppress adding a prefix route. This is for example relevant together with IFA_F_MANAGERTEMPADDR, where userspace creates autoconf generated addresses, but depending on on-link, no route for the prefix should be added. Signed-off-by:
-
David S. Miller authored
This reverts commit 12afc36e. The dependency is actually still necessary. Signed-off-by:
-
wangweidong authored
add sctp_spp_sackdelay_{enable|disable} helper function for avoiding code duplication. Signed-off-by:Wang Weidong <wangweidong1@huawei.com> Acked-by:
Neil Horman <nhorman@tuxdriver.com> Signed-off-by:
-
- 15 Jan, 2014 14 commits
-
-
Li RongQing authored
Two places defined IPV6_TCLASS_SHIFT, so we should move it into ipv6.h, and use this macro as possible. And define ip6_tclass helper to return tclass Signed-off-by:
Li RongQing <roy.qing.li@gmail.com> Signed-off-by:
-
David S. Miller authored
Sathya Perla says: ==================== be2net: patch set The following patch set is best suited for net-next as it contains code-cleanup, support for newer versions of FW cmds and a few minor fixes. Please apply. Thanks! ==================== Signed-off-by: -
Sathya Perla authored
Signed-off-by:
Sathya Perla <sathya.perla@emulex.com> Signed-off-by:
-
Suresh Reddy authored
This patch cleans-up wake-on-lan code in the following ways: 1) Removes some driver hacks in be_cmd_get_acpi_wol_cap() that were based on incorrect assumptions. 2) Uses the adapter->wol_en and wol_cap variables for checking if WoL is supported and enabled on an interface instead of referring to the exclusion list via the macro be_is_wol_supported() Signed-off-by:
Suresh Reddy <suresh.reddy@emulex.com> Signed-off-by:
-
Suresh Reddy authored
The use of NTKW_MAC_QUERY cmd has been deprecated for Skyhawk-R. Replace the last remaining usage in be_vfs_mac_query() routine. Signed-off-by:
-
Suresh Reddy authored
Instead, use the tail of the RXQ to pick the associated RXQ entry This fix is required in preparation for supporting RXQ lengths greater than 1K. For such queues, the frag index in the RX-compl entry is not valid as it is only a 10 bit entry not capable of addressing RXQs longer than 1K. Signed-off-by:
-
Suresh Reddy authored
As be2net is used even by the 40Gbps Skyhawk-R chip Signed-off-by:
-
Suresh Reddy authored
An earlier commit (f25b119c "Fix error messages while driver load for VFs") incorrectly set the adapter->cmd_privileges value for VFs (in a multi-channel config) to MAX_PRIVILEGES. This causes FW cmd failures and avoidable error logs when certian cmds are issued by a VF. Also, move the multi-channel hack to be_cmds.c inside be_cmd_get_fn_privileges() routine. Fixes: f25b119c "Fix error messages while driver load for VFs" Signed-off-by:
-
Vasundhara Volam authored
An ndo_set_mac_addr() call may be issued for a mac-addr that is already active on an interface. If so, silently ignore the request. Sending such a request to the FW, causes a "mac collision" error. The error is harmless but is avoidable noise in the kernel log. Signed-off-by:
Vasundhara Volam <vasundhara.volam@emulex.com> Signed-off-by:
-
Vasundhara Volam authored
Skyhawk-R FW does not support SET/GET_EXT_FAT_CAPABILITIES cmds via which FW logging level can be controlled. Also, the hack used in BE3 to control FW logging level via the ethtool interface is not needed in Skyhawk-R. This patch also cleans up this code by moving be_set/get_fw_log_level() routines to be_cmds.c where they belong. Signed-off-by:
-
Vasundhara Volam authored
Signed-off-by:
-
Vasundhara Volam authored
In the GET_HSW_CONFIG cmd, the "pport" field must be set only while querying the switch mode. When the "pport" field is set, the "interface_id" field must be set to the port number, otherwise, it must be set to adapter->if_handle. Signed-off-by:
-
Vasundhara Volam authored
Currently this cmd is used only for Lancer. MCC_CREATE_EXT_V1 supports larger CQ-ids and additional event codes for the async_event_bitmap field. Signed-off-by:
-
Joe Perches authored
Some systems can use the normally known u16 alignment of Ethernet addresses to save some code/text bytes and cycles. This does not change currently emitted code on x86 by gcc 4.8. Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
