Ideally the size would depend on the link speed, but the recycle
ring is created when the interface is brought up before the driver
knows the link speed. So size it for the maximum speed of a given NIC.
PowerPC is only supported on SFN7xxx and SFN8xxx NICs.
With this patch on a 40G NIC the number of calls to alloc_pages and
friends went down from about 18% to under 2%.
On a 10G NIC the number of calls to alloc_pages and friends went down
from about 15% to 0 (perf did not capture any calls during the 60
second test).
On a 100G NIC the number of calls to alloc_pages and friends went down
from about 23% to 4%.
Reported-by: Íñigo Huguet <ihuguet@redhat.com>
Signed-off-by: Martin Habets <habetsm.xilinx@gmail.com>
Link: https://lore.kernel.org/r/20220131111054.cp4f6foyinaarwbn@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The return value of kmalloc() needs to be checked.
To avoid use in efx_nic_update_stats() in case of the failure of alloc.
Fixes: b593b6f1b4 ("sfc_ef100: statistics gathering")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
dev->gso_max_segs is written under RTNL protection, or when the device is
not yet visible, but is read locklessly.
Add netif_set_gso_max_segs() helper.
Add the READ_ONCE()/WRITE_ONCE() pairs, and use netif_set_gso_max_segs()
where we can to better document what is going on.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
dev->gso_max_size is written under RTNL protection, or when the device is
not yet visible, but is read locklessly.
Add the READ_ONCE()/WRITE_ONCE() pairs, and use netif_set_gso_max_size()
where we can to better document what is going on.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert all Ethernet drivers from memcpy(... ETH_ADDR)
to eth_hw_addr_set():
@@
expression dev, np;
@@
- memcpy(dev->dev_addr, np, ETH_ALEN)
+ eth_hw_addr_set(dev, np)
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
We can treat SKB_GSO_GRE almost exactly the same as UDP tunnels, except
that we don't want to edit the outer UDP len (as there isn't one).
For SKB_GSO_GRE_CSUM, we have to use GSO_PARTIAL as the device doesn't
support offload of non-UDP outer L4 checksums.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Martin Habets <mhabets@solarflare.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
By asking the HW for the correct edits, we can make UDP tunnel TSO
work without needing GSO_PARTIAL. So don't specify it in our
netdev->gso_partial_features.
However, retain GSO_PARTIAL support, as this will be used for other
protocols later.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Martin Habets <mhabets@solarflare.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
AIUI, the NETIF_F_TSO_MANGLEID flag is a signal to the stack that a
driver may _need_ to mangle IDs in order to do TSO, and conversely
a signal from the stack that the driver is permitted to do so.
Since we support both fixed and incrementing IPIDs, we should rely
on the SKB_GSO_FIXEDID flag on a per-skb basis, rather than using
the MANGLEID feature to make all TSOs fixed-id.
Includes other minor cleanups of ef100_make_tso_desc() coding style.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The NIC only needs to know where the headers it has to edit (TCP and
inner and outer IPv4) are, which fits GSO_PARTIAL nicely.
It also supports non-PARTIAL offload of UDP tunnels, again just
needing to be told the outer transport offset so that it can edit
the UDP length field.
(It's not clear to me whether the stack will ever use the non-PARTIAL
version with the netdev feature flags we're setting here.)
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
efx_init_struct already calls this, we don't need to do it again.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
dev_close(), by way of ef100_net_stop(), already brings down the filter
table, so there's no need to do it again (which just causes lots of
WARN_ONs).
Similarly, don't bring it up ourselves, as dev_open() -> ef100_net_open()
will do it, and will fail if it's already been brought up.
Fixes: a9dc3d5612 ("sfc_ef100: RX filter table management and related gubbins")
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When downing and upping the ef100 filter table, we need to take a write
lock on efx->filter_sem, not just a read lock, because we may kfree()
the table pointers.
Without this, resets cause a WARN_ON from efx_rwsem_assert_write_locked().
Fixes: a9dc3d5612 ("sfc_ef100: RX filter table management and related gubbins")
Signed-off-by: Edward Cree <ecree@solarflare.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Actually hook up the .rx_buf_hash_valid method in EF100's nic_type.
Fixes: 068885434c ("sfc: check hash is valid before using it")
Reported-by: Martin Habets <mhabets@solarflare.com>
Signed-off-by: Edward Cree <ecree@solarflare.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The handling of the RXQ/TXQ size granularity design-params had two
problems: it had a 64-bit divide that didn't build on 32-bit platforms,
and it could divide by zero if the NIC supplied 0 as the value of the
design-param. Fix both by checking for 0 and for a granularity bigger
than our min-size; if the granularity <= EFX_MIN_DMAQ_SIZE then it fits
in 32 bits, so we can cast it to u32 for the divide.
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Edward Cree <ecree@solarflare.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
We don't yet have a .sriov_configure() to create them, though.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Self-tests for event and interrupt reception and NVRAM.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
MAC stats work much the same as on EF10, with a periodic DMA to a region
specified via an MCDI.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Bring down the TX and RX queues at ifdown, so that we can then fini the
EVQs (otherwise the MC would return EBUSY because they're still in use).
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Includes checksum offload and TSO, so declare those in our netdev features.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Several parts of the EF100 architecture are parameterised (to allow
varying capabilities on FPGAs according to resource constraints), and
these parameters are exposed to the driver through a TLV-encoded
region of the BAR.
For the most part we either don't care about these values at all or
just need to sanity-check them against the driver's assumptions, but
there are a number of TSO limits which we record so that we will be
able to check against them in the TX path when handling GSO skbs.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the future, EF100 is planned to have a credit-based scheme for
handling unsolicited events, which drivers will need to use in order
to function correctly. However, current EF100 hardware does not yet
generate unsolicited events and the credit scheme has not yet been
implemented in firmware. To prevent compatibility problems later if
the current driver is used with future firmware which does implement
it, we check for the corresponding capability flag (which that
future firmware will set), and if found, we refuse to probe.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Early in EF100 development there was a different format of event
descriptor; if the NIC is somehow running the very old firmware
which will use that format, fail the probe.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In ef100_reset(), make the MCDI call to do the reset.
Also, do a reset at start-of-day during probe, to put the function in
a clean state.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
MC_CMD_GET_CAPABILITIES now has a third word of flags; extend the
efx_has_cap() machinery to cover it.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently RX and TX-completion events are unhandled, as neither the RX
nor the TX path has been implemented yet.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Channels are probed, but actual event handling is still stubbed out.
Stub implementation of check_caps is needed because ptp.c will call into
it from efx_ptp_use_mac_tx_timestamps() to decide if it wants TXQs.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We can't actually do the MCDI to probe it fully until we have working
MCDI, which comes later, but we need efx->phy_data to be allocated so
that when we get MCDI events the link-state change handler doesn't
NULL-dereference.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We don't actually do the efx_mcdi_reset() because we don't have MCDI yet.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
No TX or RX path, no MCDI, not even an ifup/down handler.
Besides stubs, the bulk of the patch deals with reading the Xilinx
extended PCIe capability, which tells us where to find our BAR.
Though in the same module, EF100 has its own struct pci_driver,
which is named sfc_ef100.
A small number of additional nic_type methods are added; those in the
TX (tx_enqueue) and RX (rx_packet) paths are called through indirect
call wrappers to minimise the performance impact.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
