Avoid the following warnings by removing the ena_select_queue() function
and rely on the net core to do the queue selection, The issue happen
when an skb received from an interface with more queues than ena is
forwarded to the ena interface.
[ 1176.159959] eth0 selects TX queue 11, but real number of TX queues is 8
[ 1176.863976] eth0 selects TX queue 14, but real number of TX queues is 8
[ 1180.767877] eth0 selects TX queue 14, but real number of TX queues is 8
[ 1188.703742] eth0 selects TX queue 14, but real number of TX queues is 8
Fixes: 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
Signed-off-by: Kamal Heib <kheib@redhat.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is no point in initializing an ndo to NULL, therefore the
assignment is redundant and can be removed.
Signed-off-by: Kamal Heib <kheib@redhat.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Acked-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch reduces some of the lines by removing newlines
where more variables or print strings can be pushed back
to the previous line while still adhering to the styling
guidelines.
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Fail queue size calculation when the device returns maximum
TX/RX queue sizes that are smaller than the allowed minimum.
Signed-off-by: Osama Abboud <osamaabb@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The netif_* functions are used by the driver to log events into the
kernel ring (dmesg) similar to the netdev_* ones. Unlike the latter,
the netif_* function family allow the user to choose what events get
logged using ethtool:
sudo ethtool -s [interface] msglvl [msg_type] on
By default the events which get logged are slow-path related and aren't
printed often (e.g. interface up related prints). This patch removes the
NETIF_MSG_TX_DONE type (called every TX completion polling) from the
defaults and adds NETIF_MSG_IFDOWN instead as it makes more sensible
defaults.
This patch also transforms ena_down() print from netif_info into
netif_dbg (same as the analogue print in ena_up()) as it suits it
better.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Move skb_tx_timestamp() closer to the actual time the driver sends the
packets to the device.
Signed-off-by: Osama Abboud <osamaabb@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The function responsible for polling TX completions might not receive
the CPU resources it needs due to higher priority tasks running on the
requested core.
The driver might not be able to recognize such cases, but it can use its
state to suspect that they happened. If both conditions are met:
- napi hasn't been executed more than the TX completion timeout value
- napi is scheduled (meaning that we've received an interrupt)
Then it's more likely that the napi handler isn't scheduled because of
an overloaded CPU.
It was decided that for this case, the driver would wait twice as long
as the regular timeout before scheduling a reset.
The driver uses ENA_REGS_RESET_SUSPECTED_POLL_STARVATION reset reason to
indicate this case to the device.
This patch also adds more information to the ena_tx_timeout() callback.
This function is called by the kernel when it detects that a specific TX
queue has been closed for too long.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The print was re-worded to a more informative one.
Signed-off-by: Shahar Itzko <itzko@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The functionality was added to allow the drivers to create an
SQ and CQ of different sizes.
When the RX/TX SQ and CQ have the same size, such update isn't
necessary as the device can safely assume it doesn't override
unprocessed completions. However, if the SQ is larger than the CQ,
the device might "have" more completions it wants to update about
than there's room in the CQ.
There's no support for different SQ and CQ sizes, therefore,
removing the API and its usage.
'____cacheline_aligned' compiler attribute was added to
'struct ena_com_io_cq' to ensure that the removal of the
'cq_head_db_reg' field doesn't change the cache-line layout
of this struct.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Dynamic Interrupt Moderation (DIM) is a technique
designed to balance the need for timely data processing
with the desire to minimize CPU overhead.
Instead of generating an interrupt for every received
packet, the system can dynamically adjust the rate at
which interrupts are generated based on the incoming
traffic patterns.
Enabling DIM by default to improve the user experience.
DIM can be turned on/off through ethtool:
`ethtool -C <interface> adaptive-rx <on/off>`
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Osama Abboud <osamaabb@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
A few changes for better readability and style
1. Adding / Removing newlines
2. Removing an unnecessary and confusing comment
3. Using an existing variable rather than re-checking a field
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Remove io_sq->header_addr field because it is no longer
in use.
LLQ was updated to support a bounce buffer so there is
no need in saving the header address of the sq.
Signed-off-by: Nati Koler <nkoler@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Queue stats using ifconfig and ip are retrieved
via ena_get_stats64(). This function currently does not take
the xdp sent or dropped packets stats into account.
This commit adds the following xdp stats to ena_get_stats64():
tx bytes sent
tx packets sent
rx dropped packets
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-12-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The RX queue info contains information about the RX queue which might
be relevant to the kernel.
To avoid configuring this queue for different scenarios, this patch
moves the RX queue configuration to ena_up()/ena_down() function and
makes it configured every interface state toggle.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-10-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Used for better readability and debugging of XDP
flow.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-9-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This patch focuses on changes to the XDP part of the napi
polling routine.
1. Update the `napi_comp` stat only when napi is actually
complete.
2. Simplify the code by using a function pointer to the right
napi routine (XDP vs non-XDP path)
3. Remove unnecessary local variables.
4. Adjust a debug print to show the processed XDP frame index
rather than the pointer.
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-8-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This check is already done in ena_clean_rx_irq() which indirectly
calls it.
This function is called in napi context and the driver doesn't
allow to change the XDP program without performing destruction and
reinitialization of napi context (part of ena_down/ena_up sequence).
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-7-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When an XDP program is loaded the existing channels in the driver split
into two halves:
- The first half of the channels contain RX and TX rings, these queues
are used for receiving traffic and sending packets originating from
kernel.
- The second half of the channels contain only a TX ring. These queues
are used for sending packets that were redirected using XDP_TX
or XDP_REDIRECT.
Referring to the queues in the second half of the channels as "xdp_ring"
can be confusing and may give the impression that ENA has the capability
to generate an additional special queue.
This patch ensures that the xdp_ring field is exclusively used to
describe the XDP TX queue that a specific RX queue needs to utilize when
forwarding packets with XDP TX and XDP REDIRECT, preserving the
integrity of the xdp_ring field in ena_ring.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-6-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
To avoid de-referencing skb or xdp_frame when we poll for TX completion
(where they might not be in the cache), save the total TX packet size in
the ena_tx_buffer object representing the packet.
Also the 'print_once' field's type was changed from u32 to u8 to allow
adding the 'total_tx_size' without changing the total size of the
struct.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-5-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The skb and xdpf pointers cannot be set together in the driver
(each TX descriptor can send either an SKB or an XDP frame), and so it
makes more sense to put them both in a union.
This decreases the overall size of the ena_tx_buffer struct which
improves cache locality.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-4-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This change will enable the ability to use ena_xmit_common()
in functions that don't have a net_device pointer.
While it can be retrieved by dereferencing
ena_adapter (adapter->netdev), there's no reason to do it in
fast path code where this pointer is only needed for
debug prints.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-3-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
XDP system has a very large footprint in the driver's overall code.
makes the whole driver's code much harder to read.
Moving XDP code to dedicated files.
This patch doesn't make any changes to the code itself and only
cut-pastes the code into ena_xdp.c and ena_xdp.h files so the change
is purely cosmetic.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20240101190855.18739-2-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Cross-merge networking fixes after downstream PR.
Conflicts:
drivers/net/ethernet/intel/iavf/iavf_ethtool.c
3a0b5a2929fd ("iavf: Introduce new state machines for flow director")
95260816b489 ("iavf: use iavf_schedule_aq_request() helper")
https://lore.kernel.org/all/84e12519-04dc-bd80-bc34-8cf50d7898ce@intel.com/
drivers/net/ethernet/broadcom/bnxt/bnxt.c
c13e268c0768 ("bnxt_en: Fix HWTSTAMP_FILTER_ALL packet timestamp logic")
c2f8063309da ("bnxt_en: Refactor RX VLAN acceleration logic.")
a7445d69809f ("bnxt_en: Add support for new RX and TPA_START completion types for P7")
1c7fd6ee2fe4 ("bnxt_en: Rename some macros for the P5 chips")
https://lore.kernel.org/all/20231211110022.27926ad9@canb.auug.org.au/
drivers/net/ethernet/broadcom/bnxt/bnxt_ptp.c
bd6781c18cb5 ("bnxt_en: Fix wrong return value check in bnxt_close_nic()")
84793a499578 ("bnxt_en: Skip nic close/open when configuring tstamp filters")
https://lore.kernel.org/all/20231214113041.3a0c003c@canb.auug.org.au/
drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
3d7a3f2612d7 ("net/mlx5: Nack sync reset request when HotPlug is enabled")
cecf44ea1a1f ("net/mlx5: Allow sync reset flow when BF MGT interface device is present")
https://lore.kernel.org/all/20231211110328.76c925af@canb.auug.org.au/
No adjacent changes.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The get/set_rxfh ethtool ops currently takes the rxfh (RSS) parameters
as direct function arguments. This will force us to change the API (and
all drivers' functions) every time some new parameters are added.
This is part 1/2 of the fix, as suggested in [1]:
- First simplify the code by always providing a pointer to all params
(indir, key and func); the fact that some of them may be NULL seems
like a weird historic thing or a premature optimization.
It will simplify the drivers if all pointers are always present.
- Then make the functions take a dev pointer, and a pointer to a
single struct wrapping all arguments. The set_* should also take
an extack.
Link: https://lore.kernel.org/netdev/20231121152906.2dd5f487@kernel.org/ [1]
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Suggested-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Ahmed Zaki <ahmed.zaki@intel.com>
Link: https://lore.kernel.org/r/20231213003321.605376-2-ahmed.zaki@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When sending TX packets, the meta descriptor can be all zeroes
as no meta information is required (as in XDP).
This patch removes the validity check, as when
`disable_meta_caching` is enabled, such TX packets will be
dropped otherwise.
Fixes: 0e3a3f6dacf0 ("net: ena: support new LLQ acceleration mode")
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20231211062801.27891-5-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This patch fixes two issues:
Issue 1
-------
Description
```````````
Current code does not call dma_sync_single_for_cpu() to sync data from
the device side memory to the CPU side memory before the XDP code path
uses the CPU side data.
This causes the XDP code path to read the unset garbage data in the CPU
side memory, resulting in incorrect handling of the packet by XDP.
Solution
````````
1. Add a call to dma_sync_single_for_cpu() before the XDP code starts to
use the data in the CPU side memory.
2. The XDP code verdict can be XDP_PASS, in which case there is a
fallback to the non-XDP code, which also calls
dma_sync_single_for_cpu().
To avoid calling dma_sync_single_for_cpu() twice:
2.1. Put the dma_sync_single_for_cpu() in the code in such a place where
it happens before XDP and non-XDP code.
2.2. Remove the calls to dma_sync_single_for_cpu() in the non-XDP code
for the first buffer only (rx_copybreak and non-rx_copybreak
cases), since the new call that was added covers these cases.
The call to dma_sync_single_for_cpu() for the second buffer and on
stays because only the first buffer is handled by the newly added
dma_sync_single_for_cpu(). And there is no need for special
handling of the second buffer and on for the XDP path since
currently the driver supports only single buffer packets.
Issue 2
-------
Description
```````````
In case the XDP code forwarded the packet (ENA_XDP_FORWARDED),
ena_unmap_rx_buff_attrs() is called with attrs set to 0.
This means that before unmapping the buffer, the internal function
dma_unmap_page_attrs() will also call dma_sync_single_for_cpu() on
the whole buffer (not only on the data part of it).
This sync is both wasteful (since a sync was already explicitly
called before) and also causes a bug, which will be explained
using the below diagram.
The following diagram shows the flow of events causing the bug.
The order of events is (1)-(4) as shown in the diagram.
CPU side memory area
(3)convert_to_xdp_frame() initializes the
headroom with xdpf metadata
||
\/
___________________________________
| |
0 | V 4K
---------------------------------------------------------------------
| xdpf->data | other xdpf | < data > | tailroom ||...|
| | fields | | GARBAGE || |
---------------------------------------------------------------------
/\ /\
|| ||
(4)ena_unmap_rx_buff_attrs() calls (2)dma_sync_single_for_cpu()
dma_sync_single_for_cpu() on the copies data from device
whole buffer page, overwriting side to CPU side memory
the xdpf->data with GARBAGE. ||
0 4K
---------------------------------------------------------------------
| headroom | < data > | tailroom ||...|
| GARBAGE | | GARBAGE || |
---------------------------------------------------------------------
Device side memory area /\
||
(1) device writes RX packet data
After the call to ena_unmap_rx_buff_attrs() in (4), the xdpf->data
becomes corrupted, and so when it is later accessed in
ena_clean_xdp_irq()->xdp_return_frame(), it causes a page fault,
crashing the kernel.
Solution
````````
Explicitly tell ena_unmap_rx_buff_attrs() not to call
dma_sync_single_for_cpu() by passing it the ENA_DMA_ATTR_SKIP_CPU_SYNC
flag.
Fixes: f7d625adeb7b ("net: ena: Add dynamic recycling mechanism for rx buffers")
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20231211062801.27891-4-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Current xdp code drops packets larger than ENA_XDP_MAX_MTU.
This is an incorrect condition since the problem is not the
size of the packet, rather the number of buffers it contains.
This commit:
1. Identifies and drops XDP multi-buffer packets at the
beginning of the function.
2. Increases the xdp drop statistic when this drop occurs.
3. Adds a one-time print that such drops are happening to
give better indication to the user.
Fixes: 838c93dc5449 ("net: ena: implement XDP drop support")
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20231211062801.27891-3-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The ena_setup_and_create_all_xdp_queues() function freed all the
resources upon failure, after creating only xdp_num_queues queues,
instead of freeing just the created ones.
In this patch, the only resources that are freed, are the ones
allocated right before the failure occurs.
Fixes: 548c4940b9f1 ("net: ena: Implement XDP_TX action")
Signed-off-by: Shahar Itzko <itzko@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20231211062801.27891-2-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.
A suitable replacement is `strscpy` [2] due to the fact that it
guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.
host_info allocation is done in ena_com_allocate_host_info() via
dma_alloc_coherent() and is not zero initialized by alloc_etherdev_mq().
However zero initialization of the destination doesn't matter in this case,
because strscpy() guarantees a NULL termination.
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1]
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2]
Link: https://github.com/KSPP/linux/issues/90
Cc: linux-hardening@vger.kernel.org
Signed-off-by: Justin Stitt <justinstitt@google.com>
Acked-by: Arthur Kiyanovski <akiyano@amazon.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch converts some basic cases of ethtool_sprintf() to
ethtool_puts().
The conversions are used in cases where ethtool_sprintf() was being used
with just two arguments:
| ethtool_sprintf(&data, buffer[i].name);
or when it's used with format string: "%s"
| ethtool_sprintf(&data, "%s", buffer[i].name);
which both now become:
| ethtool_puts(&data, buffer[i].name);
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
xdp_do_flush_map() is deprecated and new code should use xdp_do_flush()
instead.
Replace xdp_do_flush_map() with xdp_do_flush().
Cc: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Cc: Clark Wang <xiaoning.wang@nxp.com>
Cc: Claudiu Manoil <claudiu.manoil@nxp.com>
Cc: David Arinzon <darinzon@amazon.com>
Cc: Edward Cree <ecree.xilinx@gmail.com>
Cc: Felix Fietkau <nbd@nbd.name>
Cc: Grygorii Strashko <grygorii.strashko@ti.com>
Cc: Jassi Brar <jaswinder.singh@linaro.org>
Cc: Jesse Brandeburg <jesse.brandeburg@intel.com>
Cc: John Crispin <john@phrozen.org>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: Lorenzo Bianconi <lorenzo@kernel.org>
Cc: Louis Peens <louis.peens@corigine.com>
Cc: Marcin Wojtas <mw@semihalf.com>
Cc: Mark Lee <Mark-MC.Lee@mediatek.com>
Cc: Matthias Brugger <matthias.bgg@gmail.com>
Cc: NXP Linux Team <linux-imx@nxp.com>
Cc: Noam Dagan <ndagan@amazon.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Saeed Bishara <saeedb@amazon.com>
Cc: Saeed Mahameed <saeedm@nvidia.com>
Cc: Sean Wang <sean.wang@mediatek.com>
Cc: Shay Agroskin <shayagr@amazon.com>
Cc: Shenwei Wang <shenwei.wang@nxp.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Tony Nguyen <anthony.l.nguyen@intel.com>
Cc: Vladimir Oltean <vladimir.oltean@nxp.com>
Cc: Wei Fang <wei.fang@nxp.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Arthur Kiyanovski <akiyano@amazon.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Acked-by: Martin Habets <habetsm.xilinx@gmail.com>
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Link: https://lore.kernel.org/r/20230908143215.869913-2-bigeasy@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
xdp_do_flush() should be invoked before leaving the NAPI poll function
after a XDP-redirect. This is not the case if the driver leaves via
the error path (after having a redirect in one of its previous
iterations).
Invoke xdp_do_flush() also in the error path.
Cc: Arthur Kiyanovski <akiyano@amazon.com>
Cc: David Arinzon <darinzon@amazon.com>
Cc: Noam Dagan <ndagan@amazon.com>
Cc: Saeed Bishara <saeedb@amazon.com>
Cc: Shay Agroskin <shayagr@amazon.com>
Fixes: a318c70ad152b ("net: ena: introduce XDP redirect implementation")
Acked-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
PCI core API pci_dev_id() can be used to get the BDF number for a pci
device. We don't need to compose it manually. Use pci_dev_id() to
simplify the code a little bit.
Signed-off-by: Jialin Zhang <zhangjialin11@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Shay Agroskin <shayagr@amazon.com>
Link: https://lore.kernel.org/r/20230815024248.3519068-1-zhangjialin11@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Handful of drivers currently expect to get xdp.h by virtue
of including netdevice.h. This will soon no longer be the case
so add explicit includes.
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Gerhard Engleder <gerhard@engleder-embedded.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Link: https://lore.kernel.org/r/20230803010230.1755386-2-kuba@kernel.org
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
The ENA adapters on our instances occasionally reset. Once recently
logged a UBSAN failure to console in the process:
UBSAN: shift-out-of-bounds in build/linux/drivers/net/ethernet/amazon/ena/ena_com.c:540:13
shift exponent 32 is too large for 32-bit type 'unsigned int'
CPU: 28 PID: 70012 Comm: kworker/u72:2 Kdump: loaded not tainted 5.15.117
Hardware name: Amazon EC2 c5d.9xlarge/, BIOS 1.0 10/16/2017
Workqueue: ena ena_fw_reset_device [ena]
Call Trace:
<TASK>
dump_stack_lvl+0x4a/0x63
dump_stack+0x10/0x16
ubsan_epilogue+0x9/0x36
__ubsan_handle_shift_out_of_bounds.cold+0x61/0x10e
? __const_udelay+0x43/0x50
ena_delay_exponential_backoff_us.cold+0x16/0x1e [ena]
wait_for_reset_state+0x54/0xa0 [ena]
ena_com_dev_reset+0xc8/0x110 [ena]
ena_down+0x3fe/0x480 [ena]
ena_destroy_device+0xeb/0xf0 [ena]
ena_fw_reset_device+0x30/0x50 [ena]
process_one_work+0x22b/0x3d0
worker_thread+0x4d/0x3f0
? process_one_work+0x3d0/0x3d0
kthread+0x12a/0x150
? set_kthread_struct+0x50/0x50
ret_from_fork+0x22/0x30
</TASK>
Apparently, the reset delays are getting so large they can trigger a
UBSAN panic.
Looking at the code, the current timeout is capped at 5000us. Using a
base value of 100us, the current code will overflow after (1<<29). Even
at values before 32, this function wraps around, perhaps
unintentionally.
Cap the value of the exponent used for this backoff at (1<<16) which is
larger than currently necessary, but large enough to support bigger
values in the future.
Cc: stable@vger.kernel.org
Fixes: 4bb7f4cf60e3 ("net: ena: reduce driver load time")
Signed-off-by: Krister Johansen <kjlx@templeofstupid.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Shay Agroskin <shayagr@amazon.com>
Link: https://lore.kernel.org/r/20230711013621.GE1926@templeofstupid.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The current implementation allocates page-sized rx buffers.
As traffic may consist of different types and sizes of packets,
in various cases, buffers are not fully used.
This change (Dynamic RX Buffers - DRB) uses part of the allocated rx
page needed for the incoming packet, and returns the rest of the
unused page to be used again as an rx buffer for future packets.
A threshold of 2K for unused space has been set in order to declare
whether the remainder of the page can be reused again as an rx buffer.
As a page may be reused, dma_sync_single_for_cpu() is added in order
to sync the memory to the CPU side after it was owned by the HW.
In addition, when the rx page can no longer be reused, it is being
unmapped using dma_page_unmap(), which implicitly syncs and then
unmaps the entire page. In case the kernel still handles the skbs
pointing to the previous buffers from that rx page, it may access
garbage pointers, caused by the implicit sync overwriting them.
The implicit dma sync is removed by replacing dma_page_unmap() with
dma_unmap_page_attrs() with DMA_ATTR_SKIP_CPU_SYNC flag.
The functionality is disabled for XDP traffic to avoid handling
several descriptors per packet.
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20230612121448.28829-1-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
clang 16.0.0 with W=1 reports:
drivers/net/ethernet/amazon/ena/ena_netdev.c:1901:6: error: variable 'tx_bytes' set but not used [-Werror,-Wunused-but-set-variable]
u32 tx_bytes = 0;
The variable is not used so remove it.
Signed-off-by: Simon Horman <horms@kernel.org>
Acked-by: Shay Agroskin <shayagr@amazon.com>
Link: https://lore.kernel.org/r/20230328151958.410687-1-horms@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
LLQ is auto enabled by the device and disabling it isn't supported on
new ENA generations while on old ones causes sub-optimal performance.
This patch adds advertisement of push-mode when LLQ mode is used, but
rejects an attempt to modify it.
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The ENA driver allows for two distinct values for the number of bytes
of the packet's payload that can be written directly to the device.
For a value of 224 the driver turns on Large LLQ Header mode in which
the first 224 of the packet's payload are written to the LLQ.
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
With the ability to modify LLQ entry size, the size of packet's
payload that can be written directly to the device changes.
This patch makes the driver recalculate this information every device
negotiation (also called device reset).
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Allow configuring the device with large LLQ headers. The Low Latency
Queue (LLQ) allows the driver to write the first N bytes of the packet,
along with the rest of the TX descriptors directly into device (N can be
either 96 or 224 for large LLQ headers configuration).
Having L4 TCP/UDP headers contained in the first 96 bytes of the packet
is required to get maximum performance from the device.
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move ena_calc_io_queue_size() implementation closer to the file's
beginning so that it can be later called from ena_device_init()
function without adding a function declaration.
Also add an empty line at some spots to separate logical blocks in
funcitons.
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ena nic allows xdp just if enough hw queues are available for XDP.
Take into account queues configuration setting xdp_features.
Fixes: 66c0e13ad236 ("drivers: net: turn on XDP features")
Reviewed-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A summary of the flags being set for various drivers is given below.
Note that XDP_F_REDIRECT_TARGET and XDP_F_FRAG_TARGET are features
that can be turned off and on at runtime. This means that these flags
may be set and unset under RTNL lock protection by the driver. Hence,
READ_ONCE must be used by code loading the flag value.
Also, these flags are not used for synchronization against the availability
of XDP resources on a device. It is merely a hint, and hence the read
may race with the actual teardown of XDP resources on the device. This
may change in the future, e.g. operations taking a reference on the XDP
resources of the driver, and in turn inhibiting turning off this flag.
However, for now, it can only be used as a hint to check whether device
supports becoming a redirection target.
Turn 'hw-offload' feature flag on for:
- netronome (nfp)
- netdevsim.
Turn 'native' and 'zerocopy' features flags on for:
- intel (i40e, ice, ixgbe, igc)
- mellanox (mlx5).
- stmmac
- netronome (nfp)
Turn 'native' features flags on for:
- amazon (ena)
- broadcom (bnxt)
- freescale (dpaa, dpaa2, enetc)
- funeth
- intel (igb)
- marvell (mvneta, mvpp2, octeontx2)
- mellanox (mlx4)
- mtk_eth_soc
- qlogic (qede)
- sfc
- socionext (netsec)
- ti (cpsw)
- tap
- tsnep
- veth
- xen
- virtio_net.
Turn 'basic' (tx, pass, aborted and drop) features flags on for:
- netronome (nfp)
- cavium (thunder)
- hyperv.
Turn 'redirect_target' feature flag on for:
- amanzon (ena)
- broadcom (bnxt)
- freescale (dpaa, dpaa2)
- intel (i40e, ice, igb, ixgbe)
- ti (cpsw)
- marvell (mvneta, mvpp2)
- sfc
- socionext (netsec)
- qlogic (qede)
- mellanox (mlx5)
- tap
- veth
- virtio_net
- xen
Reviewed-by: Gerhard Engleder <gerhard@engleder-embedded.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Co-developed-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Marek Majtyka <alardam@gmail.com>
Link: https://lore.kernel.org/r/3eca9fafb308462f7edb1f58e451d59209aa07eb.1675245258.git.lorenzo@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
The device supports a PCIe optimization hint, which indicates on
which NUMA the queue is currently processed. This hint is utilized
by PCIe in order to reduce its access time by accessing the
correct NUMA resources and maintaining cache coherence.
The driver calls the register update for the hint (called TPH -
TLP Processing Hint) during the NAPI loop.
Though the update is expected upon a NUMA change (when a queue
is moved from one NUMA to the other), the current logic performs
a register update when the queue is moved to a different CPU,
but the CPU is not necessarily in a different NUMA.
The changes include:
1. Performing the TPH update only when the queue has switched
a NUMA node.
2. Moving the TPH update call to be triggered only when NAPI was
scheduled from interrupt context, as opposed to a busy-polling loop.
This is due to the fact that during busy-polling, the frequency
of CPU switches for a particular queue is significantly higher,
thus, the likelihood to switch NUMA is much higher. Therefore,
providing the frequent updates to the device upon a NUMA update
are unlikely to be beneficial.
Fixes: 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RX ring can be NULL in XDP use cases where only TX queues
are configured. In this scenario, the RX interrupt moderation
value sent to the device remains in its default value of 0.
In this change, setting the default value of the RX interrupt
moderation to be the same as of the TX.
Fixes: 548c4940b9f1 ("net: ena: Implement XDP_TX action")
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make the upper bound on rx_copybreak tighter, by
making sure it is smaller than the minimum of mtu and
ENA_PAGE_SIZE. With the current upper bound of mtu,
rx_copybreak can be larger than a page. Such large
rx_copybreak will not bring any performance benefit to
the user and therefore makes no sense.
In addition, the value update was only reflected in
the adapter structure, but not applied for each ring,
causing it to not take effect.
Fixes: 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
Signed-off-by: Osama Abboud <osamaabb@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Redirecting packets with XDP Redirect is done in two phases:
1. A packet is passed by the driver to the kernel using
xdp_do_redirect().
2. After finishing polling for new packets the driver lets the kernel
know that it can now process the redirected packet using
xdp_do_flush_map().
The packets' redirection is handled in the napi context of the
queue that called xdp_do_redirect()
To avoid calling xdp_do_flush_map() each time the driver first checks
whether any packets were redirected, using
xdp_flags |= xdp_verdict;
and
if (xdp_flags & XDP_REDIRECT)
xdp_do_flush_map()
essentially treating XDP instructions as a bitmask, which isn't the case:
enum xdp_action {
XDP_ABORTED = 0,
XDP_DROP,
XDP_PASS,
XDP_TX,
XDP_REDIRECT,
};
Given the current possible values of xdp_action, the current design
doesn't have a bug (since XDP_REDIRECT = 100b), but it is still
flawed.
This patch makes the driver use a bitmask instead, to avoid future
issues.
Fixes: a318c70ad152 ("net: ena: introduce XDP redirect implementation")
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
