FW expects the driver to provide unique flow reference handles
for Tx or Rx flows. When a Tx flow and an Rx flow end up sharing
a reference handle, flow offload does not seem to work.
This could happen in the case of 2 flows having their L2 fields
wildcarded but in different direction.
Fix to incorporate the flow direction as part of the L2 key
v2: Move the dir field to the end of the bnxt_tc_l2_key struct to
fix the warning reported by kbuild test robot <lkp@intel.com>.
There is existing code that initializes the structure using
nested initializer and will warn with the new u8 field added to
the beginning. The structure also packs nicer when this new u8 is
added to the end of the structure [MChan].
Fixes: abd43a1352 ("bnxt_en: Support for 64-bit flow handle.")
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Direction of the flow is determined using src_fid. For an RX flow,
src_fid is PF's fid and for TX flow, src_fid is VF's fid. Direction
of the flow must be specified, when getting statistics for that flow.
Currently, for DECAP flow, direction is determined incorrectly, i.e.,
direction is initialized as TX for DECAP flow, instead of RX. Because
of which, stats are not reported for this DECAP flow, though it is
offloaded and there is traffic for that flow, resulting in flow age out.
This patch fixes the problem by determining the DECAP flow's direction
using correct fid. Set the flow direction in all cases for consistency
even if 64-bit flow handle is not used.
Fixes: abd43a1352 ("bnxt_en: Support for 64-bit flow handle.")
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For newly added NVM parameters, older firmware may not have the support.
Suppress the error message to avoid the unncessary error message which is
triggered when devlink calls the driver during initialization.
Fixes: 782a624d00 ("bnxt_en: Add bnxt_en initial params table and register it.")
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If FW returns FRAG_ERR in response error code, driver is resending the
command only when HWRM command returns success. Fix the code to resend
NVM_INSTALL_UPDATE command with DEFRAG install flags, if FW returns
FRAG_ERR in its response error code.
Fixes: cb4d1d6261 ("bnxt_en: Retry failed NVM_INSTALL_UPDATE with defragmentation flag enabled.")
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When both RX buffers and RX aggregation buffers have to be
replenished at the end of NAPI, post the RX aggregation buffers first
before RX buffers. Otherwise, we may run into a situation where
there are only RX buffers without RX aggregation buffers for a split
second. This will cause the hardware to abort the RX packet and
report buffer errors, which will cause unnecessary cleanup by the
driver.
Ringing the Aggregation ring doorbell first before the RX ring doorbell
will prevent some of these buffer errors. Use the same sequence during
ring initialization as well.
Fixes: 697197e5a1 ("bnxt_en: Re-structure doorbells.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
During device shutdown, the VNIC clearing sequence needs to be modified
to free the VNIC first before freeing the RSS contexts. The current
code is doing the reverse and we can get mis-directed RX completions
to CP ring ID 0 when the RSS contexts are freed and zeroed. The clearing
of RSS contexts is not required with the new sequence.
Refactor the VNIC clearing logic into a new function bnxt_clear_vnic()
and do the chip specific VNIC clearing sequence.
Fixes: 7b3af4f75b ("bnxt_en: Add RSS support for 57500 chips.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When calling debugfs functions, there is no need to ever check the
return value. The function can work or not, but the code logic should
never do something different based on this.
This cleans up a lot of unneeded code and logic around the debugfs
files, making all of this much simpler and easier to understand.
Cc: Michael Chan <michael.chan@broadcom.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use accessor functions for skb fragment's page_offset instead
of direct references, in preparation for bvec conversion.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Define the 57508, 57504, and 57502 chip IDs that are all part of the
BNXT_CHIP_P5 family of chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the new TPA feature in the 57500 chips, we need to discover the
feature first before setting up the netdev features. Refactor the
the firmware probe and init logic more cleanly into 2 functions and
and make these calls before setting up the netdev features.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Support the new expanded TPA v2 counters on 57500 B0 chips for
ethtool -S.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The new TPA implemantation has additional TPA counters that extend the
per-ring statistics block. Allocate the proper size accordingly.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current code assumes that the per ring statistics counters are
fixed. In newer chips that support a newer version of TPA, the
TPA counters are also changed. Refactor the code by defining these
counter names in arrays so that it is easy to add a new array for
a new set of counters supported by the newer chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a more optimized hardware GRO function to setup the SKB on 57500
chips. Some workaround code is no longer needed on 57500 chips and
the pseudo checksum is also calculated in hardware, so no need to
do the software pseudo checksum in the driver.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The new TPA feature on 57500 supports a larger number of concurrent TPAs
(up to 1024) divided among the functions. We need to add some logic to
map the hardware TPA ID to a software index that keeps track of each TPA
in progress. A 1:1 direct mapping without translation would be too
wasteful as we would have to allocate 1024 TPA structures for each RX
ring on each PCI function.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With all the previous refactoring, the TPA fast path can now be
modified slightly to support TPA on the new chips. The main
difference is that the agg completions are retrieved differently using
the bnxt_get_tpa_agg_p5() function on the new chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On 57500 chips, hardware GRO mode cannot be determined from the TPA
end, so we need to check bp->flags to determine if we are in hardware
GRO mode or not. Modify bnxt_set_features so that the TPA flags
in bp->flags don't change until the device is closed. This will ensure
that the fast path can safely rely on bp->flags to determine the
TPA mode.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The 2 GRO functions to set up the hardware GRO SKB fields for 2
different hardware chips have practically identical logic for
tunneled packets. Refactor the logic into a separate bnxt_gro_tunnel()
function that can be used by both functions.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On the new 57500 chips, these new RX_AGG completions are not coalesced
at the TPA_END completion. Handle these by storing them in the
array in the bnxt_tpa_info struct, as they are seen when processing
the CMPL ring.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add an aggregation array to bnxt_tpa_info struct to keep track of the
aggregation completions. The aggregation completions are not
completed at the TPA_END completion on 57500 chips so we need to
keep track of them. The array is only allocated on the new chips
when required. An agg_count field is also added to keep track of the
number of these completions.
The maximum concurrent TPA is now discovered from firmware instead of
the hardcoded 64. Add a new bp->max_tpa to keep track of maximum
configured TPA.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Refactor the TPA logic slightly, so that the code can be more easily
extended to support TPA on the new 57500 chips. In particular, the
logic to get the next aggregation completion is refactored into a
new function bnxt_get_agg() so that this operation is made more
generalized. This operation will be different on the new chip in TPA
mode. The logic to recycle the aggregation buffers has a new start
index parameter added for the same purpose.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The new chips have a slightly modified TPA interface for LRO/GRO_HW.
Modify the TPA structures so that the same structures can also be
used on the new chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Among the changes are new CoS discard counters and new ctx_hw_stats_ext
struct for the latest 5750X B0 chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While using net_dim, a dim_sample was used without ever initializing the
comps value. Added use of DIV_ROUND_DOWN_ULL() to prevent potential
overflow, it should not be a problem to save the final result in an int
because after the division by epms the value should not be larger than a
few thousand.
[ 1040.127124] UBSAN: Undefined behaviour in lib/dim/dim.c:78:23
[ 1040.130118] signed integer overflow:
[ 1040.131643] 134718714 * 100 cannot be represented in type 'int'
Fixes: 398c2b05bb ("linux/dim: Add completions count to dim_sample")
Signed-off-by: Yamin Friedman <yaminf@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Acked-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of using to_pci_dev + pci_get_drvdata,
use dev_get_drvdata to make code simpler.
Signed-off-by: Chuhong Yuan <hslester96@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation for unifying the skb_frag and bio_vec, use the fine
accessors which already exist and use skb_frag_t instead of
struct skb_frag_struct.
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Unlike legacy chips, 57500 chips don't need additional VNIC resources
for aRFS/ntuple. Fix the code accordingly so that we don't reserve
and allocate additional VNICs on 57500 chips. Without this patch,
the driver is failing to initialize when it tries to allocate extra
VNICs.
Fixes: ac33906c67 ("bnxt_en: Add support for aRFS on 57500 chips.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
kvzalloc already zeroes the memory during the allocation.
pci_alloc_consistent calls dma_alloc_coherent directly.
In commit 518a2f1925
("dma-mapping: zero memory returned from dma_alloc_*"),
dma_alloc_coherent has already zeroed the memory.
So the memset after these function is not needed.
Signed-off-by: Fuqian Huang <huangfq.daxian@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
And any other existing fields in this structure that refer to tc.
Specifically:
* tc_cls_flower_offload_flow_rule() to flow_cls_offload_flow_rule().
* TC_CLSFLOWER_* to FLOW_CLS_*.
* tc_cls_common_offload to tc_cls_common_offload.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch updates flow_block_cb_setup_simple() to use the flow block API.
Several drivers are also adjusted to use it.
This patch introduces the per-driver list of flow blocks to account for
blocks that are already in use.
Remove tc_block_offload alias.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Most drivers do the same thing to set up the flow block callbacks, this
patch adds a helper function to do this.
This preparation patch reduces the number of changes to adapt the
existing drivers to use the flow block callback API.
This new helper function takes a flow block list per-driver, which is
set to NULL until this driver list is used.
This patch also introduces the flow_block_command and
flow_block_binder_type enumerations, which are renamed to use
FLOW_BLOCK_* in follow up patches.
There are three definitions (aliases) in order to reduce the number of
updates in this patch, which go away once drivers are fully adapted to
use this flow block API.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add page_pool_destroy() in bnxt_free_rx_rings() during normal RX ring
cleanup, as Ilias has informed us that the following commit has been
merged:
1da4bbeffe ("net: core: page_pool: add user refcnt and reintroduce page_pool_destroy")
The special error handling code to call page_pool_free() can now be
removed. bnxt_free_rx_rings() will always be called during normal
shutdown or any error paths.
Fixes: 322b87ca55 ("bnxt_en: add page_pool support")
Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Cc: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Acked-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This removes contention over page allocation for XDP_REDIRECT actions by
adding page_pool support per queue for the driver. The performance for
XDP_REDIRECT actions scales linearly with the number of cores performing
redirect actions when using the page pools instead of the standard page
allocator.
v2: Fix up the error path from XDP registration, noted by Ilias Apalodimas.
Signed-off-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds basic support for XDP_REDIRECT in the bnxt_en driver. Next
patch adds the more optimized page pool support.
Signed-off-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
__bnxt_xmit_xdp() is used by XDP_TX and ethtool loopback packet transmit.
Refactor it so that it can be re-used by the XDP_REDIRECT logic.
Restructure the TX interrupt handler logic to cleanly separate XDP_TX
logic in preparation for XDP_REDIRECT.
Acked-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Renaming bnxt_xmit_xdp to __bnxt_xmit_xdp to get ready for XDP_REDIRECT
support and reduce confusion/namespace collision.
Signed-off-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some firmware versions do not support this so use the silent variant
to send the message to firmware to suppress the harmless error. This
error message is unnecessarily alarming the user.
Fixes: afdc8a8484 ("bnxt_en: Add DCBNL DSCP application protocol support.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In an earlier commit to improve NQ reservations on 57500 chips, we
set the resv_irqs on the 57500 VFs to the fixed value assigned by
the PF regardless of how many are actually used. The current
code assumes that resv_irqs minus the ones used by the network driver
must be the ones for the RDMA driver. This is no longer true and
we may return more MSIX vectors than requested, causing inconsistency.
Fix it by capping the value.
Fixes: 01989c6b69 ("bnxt_en: Improve NQ reservations.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current logic assumes that the RDMA driver uses one statistics
context adjacent to the ones used by the network driver. This
assumption is not true and the statistics context used by the
RDMA driver is tied to its MSIX base vector. This wrong assumption
can cause RDMA driver failure after changing ethtool rings on the
network side. Fix the statistics reservation logic accordingly.
Fixes: 780baad44f ("bnxt_en: Reserve 1 stat_ctx for RDMA driver.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After ethtool loopback packet tests, we re-open the nic for the next
IRQ test. If the open fails, we must not proceed with the IRQ test
or we will crash with NULL pointer dereference. Fix it by checking
the bnxt_open_nic() return code before proceeding.
Reported-by: Somasundaram Krishnasamy <somasundaram.krishnasamy@oracle.com>
Fixes: 67fea463fd ("bnxt_en: Add interrupt test to ethtool -t selftest.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some chips with older firmware can continue to perform DMA read from
context memory even after the memory has been freed. In the PCI shutdown
method, we need to call pci_disable_device() to shutdown DMA to prevent
this DMA before we put the device into D3hot. DMA memory request in
D3hot state will generate PCI fatal error. Similarly, in the driver
remove method, the context memory should only be freed after DMA has
been shutdown for correctness.
Fixes: 98f04cf0f1 ("bnxt_en: Check context memory requirements from firmware.")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Saeed Mamameed says:
====================
Generic DIM
From: Tal Gilboa and Yamin Fridman
Implement net DIM over a generic DIM library, add RDMA DIM
dim.h lib exposes an implementation of the DIM algorithm for
dynamically-tuned interrupt moderation for networking interfaces.
We want a similar functionality for other protocols, which might need to
optimize interrupts differently. Main motivation here is DIM for NVMf
storage protocol.
Current DIM implementation prioritizes reducing interrupt overhead over
latency. Also, in order to reduce DIM's own overhead, the algorithm might
take some time to identify it needs to change profiles. While this is
acceptable for networking, it might not work well on other scenarios.
Here we propose a new structure to DIM. The idea is to allow a slightly
modified functionality without the risk of breaking Net DIM behavior for
netdev. We verified there are no degradations in current DIM behavior with
the modified solution.
Suggested solution:
- Common logic is implemented in lib/dim/dim.c
- Net DIM (existing) logic is implemented in lib/dim/net_dim.c, which uses
the common logic in dim.c
- Any new DIM logic will be implemented in "lib/dim/new_dim.c".
This new implementation will expose modified versions of profiles,
dim_step() and dim_decision().
- DIM API is declared in include/linux/dim.h for all implementations.
Pros for this solution are:
- Zero impact on existing net_dim implementation and usage
- Relatively more code reuse (compared to two separate solutions)
- Increased extensibility
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Moved all logic from dim.h and net_dim.h to dim.c and net_dim.c.
This is both more structurally appealing and would allow to only
expose externally used functions.
Signed-off-by: Tal Gilboa <talgi@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Removed 'net' prefix from functions and structs used by external drivers.
Signed-off-by: Tal Gilboa <talgi@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In order to avoid confusion between the function and the similarly
named struct.
In preparation for removing the 'net' prefix from dim members.
Signed-off-by: Tal Gilboa <talgi@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
