| Age | Commit message (Collapse) | Author |
|---|
|
1. pcaps are free'd right after AQ routines are done, no need for
devm_'s
2. a test frame for loopback test in ethtool -t is destroyed at the end
of the test so we don't need devm_ here either.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
ice_down() clears QINT_TQCTL_CAUSE_ENA_M bit twice, which is not
necessary. First clearing happens in ice_vsi_dis_irq() and second in
ice_vsi_stop_tx_ring() - remove the first one.
While at it, make ice_vsi_dis_irq() static as ice_down() is the only
current caller of it.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
The indentation for ice_set_ctx and ice_write_rxq_ctx breaks the function
name after the return type. This style of breaking is used a lot throughout
the ice driver, even in cases where its not actually helpful for
readability. We no longer prefer this style of line splitting in the
driver, and new code is avoiding it.
Normally, I would leave this alone unless the actual function contents or
description needed updating. However, a future change is going to add
inverse functions for converting packed context to unpacked context
structures. To keep this code uniform with the existing set functions, fix
up the style to the modern format of keeping the type on the same line.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
The functions used to pack the Tx and Rx context into the hardware format
rely on using BIT() and then subtracting 1 to get a bitmask. These
functions even have a comment about how x86 machines can't use this method
for certain widths because the SHL instructions will not work properly.
The Linux kernel already provides the GENMASK macro for generating a
suitable bitmask. Further, GENMASK is capable of generating the mask
including the shift_width. Since width is the total field width, take care
to subtract one to get the final bit position.
Since we now include the shifted bits as part of the mask, shift the source
value first before applying the mask.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
In ice_common.c there are 4 functions used for converting the unpacked
software Tx and Rx context structure data into the packed format used by
hardware. These functions have extremely generic names:
* ice_write_byte
* ice_write_word
* ice_write_dword
* ice_write_qword
When I saw these function names my first thought was "write what? to
where?". Understanding what these functions do requires looking at the
implementation details. The functions take bits from an unpacked structure
and copy them into the packed layout used by hardware.
As part of live migration, we will want functions which perform the inverse
operation of reading bits from the packed layout and copying them into the
unpacked format. Naming these as "ice_read_byte", etc would be very
confusing since they appear to write data.
In preparation for adding this new inverse operation, rename the existing
functions to use the prefix "ice_pack_ctx_". This makes it clear that they
perform the bit packing while copying from the unpacked software context
structure to the packed hardware context.
The inverse operations can then neatly be named ice_unpack_ctx_*, clearly
indicating they perform the bit unpacking while copying from the packed
hardware context to the unpacked software context structure.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
The lan_vsi_num field of the VF structure is no longer used for any
purpose. Remove it.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
When initializing over virtchnl, the PF is required to pass a VSI ID to the
VF as part of its capabilities exchange. The VF driver reports this value
back to the PF in a variety of commands. The PF driver validates that this
value matches the value it sent to the VF.
Some hardware families such as the E700 series could use this value when
reading RSS registers or communicating directly with firmware over the
Admin Queue.
However, E800 series hardware does not support any of these interfaces and
the VF's only use for this value is to report it back to the PF. Thus,
there is no requirement that this value be an actual VSI ID value of any
kind.
The PF driver already does not trust that the VF sends it a real VSI ID.
The VSI structure is always looked up from the VF structure. The PF does
validate that the VSI ID provided matches a VSI associated with the VF, but
otherwise does not use the VSI ID for any purpose.
Instead of reporting the VSI number relative to the PF space, report a
fixed value of 1. When communicating with the VF over virtchnl, validate
that the VSI number is returned appropriately.
This avoids leaking information about the firmware of the PF state.
Currently the ice driver only supplies a VF with a single VSI. However, it
appears that virtchnl has some support for allowing multiple VSIs. I did
not attempt to implement this. However, space is left open to allow further
relative indexes if additional VSIs are provided in future feature
development. For this reason, keep the ice_vc_isvalid_vsi_id function in
place to allow extending it for multiple VSIs in the future.
This change will also simplify handling of live migration in a future
series. Since we no longer will provide a real VSI number to the VF, there
will be no need to keep track of this number when migrating to a new host.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
The ice_vc_fdir_param_check() function validates that the VSI ID of the
virtchnl flow director command matches the VSI number of the VF. This is
already checked by the call to ice_vc_isvalid_vsi_id() immediately
following this.
This check is unnecessary since ice_vc_isvalid_vsi_id() already confirms
this by checking that the VSI ID can locate the VSI associated with the VF
structure.
Furthermore, a following change is going to refactor the ice driver to
report VSI IDs using a relative index for each VF instead of reporting the
PF VSI number. This additional check would break that logic since it
enforces that the VSI ID matches the VSI number.
Since this check duplicates the logic in ice_vc_isvalid_vsi_id() and gets
in the way of refactoring that logic, remove it.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
The ice_vc_isvalid_q_id() function takes a VSI index and a queue ID. It
looks up the VSI from its index, and then validates that the queue number
is valid for that VSI.
The VSI ID passed is typically a VSI index from the VF. This VSI number is
validated by the PF to ensure that it matches the VSI associated with the
VF already.
In every flow where ice_vc_isvalid_q_id() is called, the PF driver already
has a pointer to the VSI associated with the VF. This pointer is obtained
using ice_get_vf_vsi(), rather than looking up the VSI using the index sent
by the VF.
Since we already know which VSI to operate on, we can modify
ice_vc_isvalid_q_id() to take a VSI pointer instead of a VSI index. Pass
the VSI we found from ice_get_vf_vsi() instead of re-doing the lookup. This
removes some unnecessary computation and scanning of the VSI list.
It also removes the last place where the driver directly used the VSI
number from the VF. This will pave the way for refactoring to communicate
relative VSI numbers to the VF instead of absolute numbers from the PF
space.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
This error message is at best not really helpful and at worst
misleading. If we're here in idpf_intr_rel we're likely trying to do
remove or reset. If we're in reset, this message will fail because we
lose the virtchnl on reset and HW is going to clean up those resources
regardless in that case. If we're in remove and we get an error here,
we're going to reset the device at the end of remove anyway so not a big
deal. Just remove this message it's not useful.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
While we're here improving virtchnl we can include two minor fixes for
the lower level ctrlq flow.
This adds a memory barrier to idpf_post_rx_buffs before we update tail
on the controlq. We should make sure our writes have had a chance to
finish before we tell HW it can touch them.
This also removes some defensive programming in idpf_ctrlq_recv. The
caller should not be using a num_q_msg value of zero or more than the
ring size and it's their responsibility to call functions sanely.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
In idpf_remove we need to tear down the virtchnl core with
idpf_vc_core_deinit so we can free up resources and leave things in a
good state. However, in the case where we failed to establish VC
communications we may not have ever actually successfully initialized
the virtchnl core.
This fixes it by setting a bit once we successfully init the virtchnl
core. Then, in deinit, we'll check for it before going on further,
otherwise we just return. Also clear the bit at the end of deinit so we
know it's gone now.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
We can now remove a bunch of gross code we don't need anymore like the
vc state bits and vc_buf_lock since everything is using transaction API
now.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Igor Bagnucki <igor.bagnucki@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
Now that all the messages are using the transaction API, we can rework
idpf_recv_mb_msg quite a lot to simplify it. Due to this, we remove
idpf_find_vport as no longer used and alter idpf_recv_event_msg
slightly.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
There are situations where the driver needs to add a MAC filter but
we're explicitly not allowed to sleep so we can wait for a virtchnl
message to complete.
This adds an async_handler for asynchronously sent messages for MAC
filters so that we can better handle if there's an error of some kind.
If success we don't need to do anything else, but if we failed to
program the new filter we really should remove it from our list of MAC
filters. If we don't remove bad filters, what I expect to happen is
after a reset of some kind we try to program the MAC filter again and it
fails again. This is clearly wrong and I would expect to be confusing
for the user.
It could also be the failure is for a delete MAC filter message but
those filters get deleted regardless. Not much we can do about a delete
failure.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
This takes care of RSS/SRIOV/MAC and other misc virtchnl messages. This
again is mostly mechanical.
In absence of an async_handler for MAC filters, this will simply
generically report any errors from idpf_vc_xn_forward_async. This
maintains the existing behavior. Follow up patch will add an async
handler for MAC filters to remove bad filters from our list.
While we're here we can also make the code much nicer by converting some
variables to auto-variables where appropriate. This makes it cleaner and
less prone to memory leaking.
There's still a bit more cleanup we can do here to remove stuff that's
not being used anymore now; follow-up patches will take care of loose
ends.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
This reworks queue specific virtchnl messages to use the added
transaction API. It is fairly mechanical and generally makes the
functions using it more simple. Functions using transaction API no
longer need to take the vc_buf_lock since it's not using it anymore.
After filling out an idpf_vc_xn_params struct, idpf_vc_xn_exec takes
care of the send and recv handling.
This also converts those functions where appropriate to use
auto-variables instead of manually calling kfree. This greatly
simplifies the memory alloc paths and makes them less prone memory
leaks.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Igor Bagnucki <igor.bagnucki@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
This reworks the way vport related virtchnl messages work to take
advantage of the added transaction API. It is fairly mechanical as, to
use the transaction API, the function just needs to fill out an
appropriate idpf_vc_xn_params struct to pass to idpf_vc_xn_exec which
will take care of the actual send and recv.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Igor Bagnucki <igor.bagnucki@intel.com>
Co-developed-by: Joshua Hay <joshua.a.hay@intel.com>
Signed-off-by: Joshua Hay <joshua.a.hay@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
This starts refactoring how virtchnl messages are handled by adding a
transaction manager (idpf_vc_xn_manager).
There are two primary motivations here which are to enable handling of
multiple messages at once and to make it more robust in general. As it
is right now, the driver may only have one pending message at a time and
there's no guarantee that the response we receive was actually intended
for the message we sent prior.
This works by utilizing a "cookie" field of the message descriptor. It
is arbitrary what data we put in the cookie and the response is required
to have the same cookie the original message was sent with. Then using a
"transaction" abstraction that uses the completion API to pair responses
to the message it belongs to.
The cookie works such that the first half is the index to the
transaction in our array, and the second half is a "salt" that gets
incremented every message. This enables quick lookups into the array and
also ensuring we have the correct message. The salt is necessary because
after, for example, a message times out and we deem the response was
lost for some reason, we could theoretically reuse the same index but
using a different salt ensures that when we do actually get a response
it's not the old message that timed out previously finally coming in.
Since the number of transactions allocated is U8_MAX and the salt is 8
bits, we can never have a conflict because we can't roll over the salt
without using more transactions than we have available.
This starts by only converting the VIRTCHNL2_OP_VERSION message to use
this new transaction API. Follow up patches will convert all virtchnl
messages to use the API.
Tested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Igor Bagnucki <igor.bagnucki@intel.com>
Co-developed-by: Joshua Hay <joshua.a.hay@intel.com>
Signed-off-by: Joshua Hay <joshua.a.hay@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
idpf.h is quite heavy. We can reduce the burden a fair bit by
introducing an idpf_virtchnl.h file. This mostly just moves function
declarations but there are many of them. This also makes an attempt to
group those declarations in a way that makes some sense instead of
mishmashed.
Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
Pass the link conf to the abort_channel_switch driver
method so the driver can handle things correctly.
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Reviewed-by: Miriam Rachel Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://msgid.link/20240228095718.27f621106ddd.Iadd3d69b722ffe5934779a32a0e4e596a4e33ed4@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
|
|
In MLO, we need the link id in the GTK key to be given by
the driver after rekeying in wowlan, so add that.
Signed-off-by: Shaul Triebitz <shaul.triebitz@intel.com>
Reviewed-by: Miriam Rachel Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://msgid.link/20240228094500.ce1bfc83a680.I43a6f8ab2804ee07116a37d5b9ec601b843464b1@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
|
|
A few models require *.pnvm files while we don't declare them via
MODULE_FIRMWARE(). This resulted in the breakage of WiFi on the
system that relies on the information from modinfo (e.g. openSUSE
installer image).
This patch adds those missing MODULE_FIRMWARE() entries for *.pnvm
files.
type=feature
ticket=none
Link: https://bugzilla.opensuse.org/show_bug.cgi?id=1207553
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/20240228163837.4320-1-tiwai@suse.de
[move to appropriate files]
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
|
|
The IPA platform device is now only used as the structure containing
the IPA device structure. Replace the platform device pointer with
a pointer to the device structure.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rather than using the platform device pointer field in the IPA
pointer, pass a platform device pointer to ipa_smp2p_init(). Use
that pointer throughout that function.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rather than using the platform device pointer field in the IPA
pointer, pass a platform device pointer to ipa_smp2p_irq_init().
Use that pointer throughout that function (without assuming it's
the same as the IPA platform device pointer).
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rather than using the platform device pointer field in the IPA
pointer, pass a platform device pointer to ipa_mem_init(). Use
that pointer throughout that function.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rather than using the platform device pointer field in the IPA
pointer, pass a platform device pointer to ipa_reg_init(). Use
that pointer throughout that function.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Create a new function ipa_interrupt_init() that is called at probe
time to allocate and initialize the IPA interrupt data structure.
Create ipa_interrupt_exit() as its inverse.
This follows the normal IPA driver pattern of *_init() functions
doing things that can be done before access to hardware is required.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Change the return type of ipa_interrupt_config() to be an error
code rather than an IPA interrupt structure pointer, and assign the
the pointer within that function.
Change ipa_interrupt_deconfig() to take the IPA pointer as argument
and have it invalidate the ipa->interrupt pointer.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Do not set rtnl_link_stats64 fields to zero, since they are zeroed
before ops->ndo_get_stats64 is called in core dev_get_stats() function.
Also, simplify the data collection by removing the temporary variable.
Signed-off-by: Breno Leitao <leitao@debian.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
With commit 34d21de99cea9 ("net: Move {l,t,d}stats allocation to core and
convert veth & vrf"), stats allocation could be done on net core
instead of this driver.
With this new approach, the driver doesn't have to bother with error
handling (allocation failure checking, making sure free happens in the
right spot, etc). This is core responsibility now.
Remove the allocation in the nlmon driver and leverage the network
core allocation.
Signed-off-by: Breno Leitao <leitao@debian.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The last patch which added support to extend the firmware shared
data to add channel data information has introduced a bug due to
the reserved space not adjusted accordingly.
This patch fixes the issue and also adds BUILD_BUG to avoid this
regression error.
Fixes: 997814491cee ("Octeontx2-af: Fetch MAC channel info from firmware")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
struct net_device poll_dev in struct igc_q_vector was added
in one of the initial commits, but never used.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
To record the stats of header split packets, three stats are added in
the driver's ethtool stats.
- rx_hsplit_pkt is the split packets count with header split
- rx_hsplit_bytes is the received header bytes count with header split
- rx_hsplit_unsplit_pkt is the unsplit packet count due to header buffer
overflow or zero header length when header split is enabled
Currently, it's entering the stats_update critical section more than
once per packet. We have plans to avoid that in the future change to let
all the stats_update happen in one place at the end of
`gve_rx_poll_dqo`.
Co-developed-by: Ziwei Xiao <ziweixiao@google.com>
Signed-off-by: Ziwei Xiao <ziweixiao@google.com>
Signed-off-by: Jeroen de Borst <jeroendb@google.com>
Reviewed-by: Praveen Kaligineedi <pkaligineedi@google.com>
Reviewed-by: Harshitha Ramamurthy <hramamurthy@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add header buffers and ethtool support to enable header split via the
tcp-data-split flag in ethtool's ringparam config. A coherent dma memory
is allocated for the header buffers. There is one header buffer per ring
entry by calculating the offset to the header-buffers starting address.
The header buffer is always copied directly into the skb and payload is
always added as frags. When there is a header buffer overflow or the
header length is 0, the driver places the whole unsplit packet in frags.
When toggling header split, the driver will call gve_adjust_config to
set its queues appropriately. If header split is enabled by the user and
the max packet buffer size is no less than 4KB, driver will set the
packet buffer size as 4KB to support TCP_ZEROCOPY_RECEIVE. Otherwise the
driver will use the default 2KB as the packet buffer size.
`ethtool -G <dev> tcp-data-split on/off` is the command to toggle header
split.
`ethtool -g <dev>` will show the status of header split with the field
of `tcp-data-split`.
Co-developed-by: Ziwei Xiao <ziweixiao@google.com>
Signed-off-by: Ziwei Xiao <ziweixiao@google.com>
Signed-off-by: Jeroen de Borst <jeroendb@google.com>
Reviewed-by: Praveen Kaligineedi <pkaligineedi@google.com>
Reviewed-by: Harshitha Ramamurthy <hramamurthy@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
To enable header split via ethtool, we first need to query the device to
get the max rx buffer size and header buffer size. Add a device option
to get these values and store them in the driver. If the header buffer
size received from the device is non-zero, it means header split is
supported in the device.
Currently the max rx buffer size will only be used when header split is
enabled which will set the data_buffer_size_dqo to be the max rx buffer
size. Also change the data_buffer_size_dqo from int to u16 since we are
modifying it and making it to be consistent with max_rx_buffer_size.
Co-developed-by: Ziwei Xiao <ziweixiao@google.com>
Signed-off-by: Ziwei Xiao <ziweixiao@google.com>
Signed-off-by: Jeroen de Borst <jeroendb@google.com>
Reviewed-by: Praveen Kaligineedi <pkaligineedi@google.com>
Reviewed-by: Harshitha Ramamurthy <hramamurthy@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
syzbot triggered a bug in geneve_rx() [1]
Issue is similar to the one I fixed in commit 8d975c15c0cd
("ip6_tunnel: make sure to pull inner header in __ip6_tnl_rcv()")
We have to save skb->network_header in a temporary variable
in order to be able to recompute the network_header pointer
after a pskb_inet_may_pull() call.
pskb_inet_may_pull() makes sure the needed headers are in skb->head.
[1]
BUG: KMSAN: uninit-value in IP_ECN_decapsulate include/net/inet_ecn.h:302 [inline]
BUG: KMSAN: uninit-value in geneve_rx drivers/net/geneve.c:279 [inline]
BUG: KMSAN: uninit-value in geneve_udp_encap_recv+0x36f9/0x3c10 drivers/net/geneve.c:391
IP_ECN_decapsulate include/net/inet_ecn.h:302 [inline]
geneve_rx drivers/net/geneve.c:279 [inline]
geneve_udp_encap_recv+0x36f9/0x3c10 drivers/net/geneve.c:391
udp_queue_rcv_one_skb+0x1d39/0x1f20 net/ipv4/udp.c:2108
udp_queue_rcv_skb+0x6ae/0x6e0 net/ipv4/udp.c:2186
udp_unicast_rcv_skb+0x184/0x4b0 net/ipv4/udp.c:2346
__udp4_lib_rcv+0x1c6b/0x3010 net/ipv4/udp.c:2422
udp_rcv+0x7d/0xa0 net/ipv4/udp.c:2604
ip_protocol_deliver_rcu+0x264/0x1300 net/ipv4/ip_input.c:205
ip_local_deliver_finish+0x2b8/0x440 net/ipv4/ip_input.c:233
NF_HOOK include/linux/netfilter.h:314 [inline]
ip_local_deliver+0x21f/0x490 net/ipv4/ip_input.c:254
dst_input include/net/dst.h:461 [inline]
ip_rcv_finish net/ipv4/ip_input.c:449 [inline]
NF_HOOK include/linux/netfilter.h:314 [inline]
ip_rcv+0x46f/0x760 net/ipv4/ip_input.c:569
__netif_receive_skb_one_core net/core/dev.c:5534 [inline]
__netif_receive_skb+0x1a6/0x5a0 net/core/dev.c:5648
process_backlog+0x480/0x8b0 net/core/dev.c:5976
__napi_poll+0xe3/0x980 net/core/dev.c:6576
napi_poll net/core/dev.c:6645 [inline]
net_rx_action+0x8b8/0x1870 net/core/dev.c:6778
__do_softirq+0x1b7/0x7c5 kernel/softirq.c:553
do_softirq+0x9a/0xf0 kernel/softirq.c:454
__local_bh_enable_ip+0x9b/0xa0 kernel/softirq.c:381
local_bh_enable include/linux/bottom_half.h:33 [inline]
rcu_read_unlock_bh include/linux/rcupdate.h:820 [inline]
__dev_queue_xmit+0x2768/0x51c0 net/core/dev.c:4378
dev_queue_xmit include/linux/netdevice.h:3171 [inline]
packet_xmit+0x9c/0x6b0 net/packet/af_packet.c:276
packet_snd net/packet/af_packet.c:3081 [inline]
packet_sendmsg+0x8aef/0x9f10 net/packet/af_packet.c:3113
sock_sendmsg_nosec net/socket.c:730 [inline]
__sock_sendmsg net/socket.c:745 [inline]
__sys_sendto+0x735/0xa10 net/socket.c:2191
__do_sys_sendto net/socket.c:2203 [inline]
__se_sys_sendto net/socket.c:2199 [inline]
__x64_sys_sendto+0x125/0x1c0 net/socket.c:2199
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xcf/0x1e0 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x63/0x6b
Uninit was created at:
slab_post_alloc_hook mm/slub.c:3819 [inline]
slab_alloc_node mm/slub.c:3860 [inline]
kmem_cache_alloc_node+0x5cb/0xbc0 mm/slub.c:3903
kmalloc_reserve+0x13d/0x4a0 net/core/skbuff.c:560
__alloc_skb+0x352/0x790 net/core/skbuff.c:651
alloc_skb include/linux/skbuff.h:1296 [inline]
alloc_skb_with_frags+0xc8/0xbd0 net/core/skbuff.c:6394
sock_alloc_send_pskb+0xa80/0xbf0 net/core/sock.c:2783
packet_alloc_skb net/packet/af_packet.c:2930 [inline]
packet_snd net/packet/af_packet.c:3024 [inline]
packet_sendmsg+0x70c2/0x9f10 net/packet/af_packet.c:3113
sock_sendmsg_nosec net/socket.c:730 [inline]
__sock_sendmsg net/socket.c:745 [inline]
__sys_sendto+0x735/0xa10 net/socket.c:2191
__do_sys_sendto net/socket.c:2203 [inline]
__se_sys_sendto net/socket.c:2199 [inline]
__x64_sys_sendto+0x125/0x1c0 net/socket.c:2199
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xcf/0x1e0 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x63/0x6b
Fixes: 2d07dc79fe04 ("geneve: add initial netdev driver for GENEVE tunnels")
Reported-and-tested-by: syzbot+6a1423ff3f97159aae64@syzkaller.appspotmail.com
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The MODULE_AUTHOR macro is supposed to be a person
not a company.
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Clean up complaints from an xmastree.py scan.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use the kernel's CQE dim table to align better with the
driver's use of completion queues, and use the tx moderation
when using Tx interrupts.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
An earlier change moved the hwstamp queue check into a helper
function with an unlikely(). However, it makes more sense for
the caller to decide if it's likely() or unlikely(), so make
the change to support that.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
To help make sure we're only accessing things we really need
to access we can cut down on the q->lif->netdev references by
using q->dev which is already in cache.
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Instead of using q->lif->netdev, just pass the netdev when it's
locally defined.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If there is a lot of transmit traffic the driver can get into a
situation that the device is starved due to the doorbell never
being rung. This can happen if xmit_more is set constantly
and __netdev_tx_sent_queue() keeps returning false. Fix this
by checking if the queue needs to be stopped right before
calling __netdev_tx_sent_queue(). Use MAX_SKB_FRAGS + 1 as the
stop condition because that's the maximum number of frags
supported for non-TSO transmit.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver currently calls netdev_tx_completed_queue() for every
Tx completion. However, this API is only meant to be called once
per NAPI if any Tx work is done. Make the necessary changes to
support calling netdev_tx_completed_queue() only once per NAPI.
Also, use the __netdev_tx_sent_queue() API, which supports the
xmit_more functionality.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Make use of napi_consume_skb so that skb recycling
can happen by way of the napi_skb_cache.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Perf was showing some hot spots in ionic_tx_descs_needed()
for TSO traffic. Rework the function to return sooner where
possible.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Cut down the number of default Tx and Rx descriptors to save
initial memory requirements.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Currently the driver attempts to wake the Tx queue
for every descriptor processed. However, this is
overkill and can cause thrashing since Tx xmit can be
running concurrently on a different CPU than Tx clean.
Fix this by refactoring Tx cq servicing into its own
function so the Tx wake code can run after processing
all Tx descriptors.
The driver isn't using the expected memory barriers
to make sure the stop/start bits are coherent. Fix
this by making sure to use the correct memory barriers.
Also, the driver is using the wake API during Tx
xmit even though it's already scheduled. Fix this by
using the start API during Tx xmit.
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
