A GSI event ring causes an event control interrupt to fire whenever
its state changes (between NOT_ALLOCATED and ALLOCATED). No event
ring should ever change state except when we request it to.
Currently, we permit *all* events rings to generate event control
interrupts--even those that are never used. And we enable event
control interrupts essentially at all times, from setup to teardown.
Instead, only enable the event control interrupt type for the
duration of an event ring command, and when doing so, only allow
the event ring being operated upon to cause the interrupt to fire.
Disallow all event rings from issuing the event control interrupt
in gsi_irq_setup().
Because an event ring's interrupt is only enabled when needed,
there is no longer any need to zero the event channel mask in
gsi_irq_disable().
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A GSI channel causes a channel control interrupt to fire whenever
its state changes (between NOT_ALLOCATED, ALLOCATED, STARTED, etc.).
We do not support inter-EE channel commands (initiated by other EEs),
so no channel should ever change state except when we request it to.
Currently, we permit *all* channels to generate channel control
interrupts--even those that are never used. And we enable channel
control interrupts essentially at all times, from setup to teardown.
Instead, disable all channel control interrupts initially in
gsi_irq_setup(), and only enable the channel control interrupt
type for the duration of a channel command. When doing so, only
allow the channel being operated upon to cause the interrupt to
fire.
Because a channel's interrupt is now enabled only when needed (one
channel at a time), there is no longer any need to zero the channel
mask in gsi_irq_disable().
Add new gsi_irq_type_enable() and gsi_irq_type_disable() as helper
functions to control whether a given GSI interrupt type is enabled.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Keep track of the set of GSI interrupt types that are currently
enabled by recording the mask value to write (or last written) to
the TYPE_IRQ_MSK register.
Create a new helper function gsi_irq_type_update() to handle
actually writing the register.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Introduce gsi_irq_setup() and gsi_irq_teardown() to disable all
GSI interrupts when first setting up GSI hardware, and to clean
things up when we're done.
Re-enable all GSI interrupt types in gsi_irq_enable(), but do
so only after each of the type-specific interrupt masks has
been configured. Similarly, disable all interrupt types in
gsi_irq_disable()--first--before zeroing out the type-specific
masks.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Define the GSI interrupt types with an enumerated type whose values
are the bit positions representing each interrupt type. Include a
short comment describing how each interrupt type is used.
Build up the enabled interrupt mask explicitly in gsi_irq_enable(),
and get rid of the definition of GSI_CNTXT_TYPE_IRQ_MSK_ALL.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Rename the "event_enable_bitmap" field of the GSI structure to be
"ieob_enabled_bitmap". An upcoming patch will cache the last value
stored for another interrupt mask and this is a more direct naming
convention to follow.
Add a few comments to explain the bitmap fields in the GSI structure.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Introduce gsi_irq_init() and gsi_irq_exit(), to encapsulate looking
up the GSI IRQ and registering its handler. Call gsi_irq_init() a
little later in gsi_init(), and initialize the completion earlier.
The IRQ handler accesses both the GSI virtual memory pointer and the
completion, and this way these things will have been initialized
before the gsi_irq() can ever be called.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The GSI code is now exposed to IPA version numbers, and we handle
version-specific behavior based on the IPA version.
Modify some comments that talk about GSI versions so they reference
IPA versions instead. Correct version number errors in a couple of
these comments.
The (comment) mapping between IPA and GSI versions in the definition
of the ipa_version enumerated type remains.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This driver transports LAPB (X.25 link layer) frames over TTY links.
I can safely say that this driver has no actual user because it was
not working at all until:
commit 8fdcabeac3 ("drivers/net/wan/x25_asy: Fix to make it work")
The code in its current state still has problems:
1.
The uses of "struct x25_asy" in x25_asy_unesc (when receiving) and in
x25_asy_write_wakeup (when sending) are not protected by locks against
x25_asy_change_mtu's changing of the transmitting/receiving buffers.
Also, all "netif_running" checks in this driver are not protected by
locks against the ndo_stop function.
2.
The driver stops all TTY read/write when the netif is down.
I think this is not right because this may cause the last outgoing frame
before the netif goes down to be incompletely transmitted, and the first
incoming frame after the netif goes up to be incompletely received.
And there may also be other problems.
I was planning to fix these problems but after recent discussions about
deleting other old networking code, I think we may just delete this
driver, too.
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Acked-by: Martin Schiller <ms@dev.tdt.de>
Link: https://lore.kernel.org/r/20201105073434.429307-1-xie.he.0141@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Radhey Shyam Pandey says:
====================
net: axienet: Dynamically enable MDIO interface
This patchset dynamically enable MDIO interface. The background for this
change is coming from Cadence GEM controller(macb) in which MDC is active
only during MDIO read or write operations while the PHY registers are
read or written. It is implemented as an IP feature.
For axiethernet as dynamic MDC enable/disable is not supported in hw
we are implementing it in sw. This change doesn't affect any existing
functionality.
====================
Link: https://lore.kernel.org/r/1604402770-78045-1-git-send-email-radhey.shyam.pandey@xilinx.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
MDIO spec does not require an MDC at all times, only when MDIO
transactions are occurring. This patch allows the xilinx_axienet
driver to disable the MDC when not in use, and re-enable it when
needed. It also simplifies the driver by removing MDC disable
and enable in device reset sequence.
Signed-off-by: Clayton Rayment <clayton.rayment@xilinx.com>
Signed-off-by: Radhey Shyam Pandey <radhey.shyam.pandey@xilinx.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Introduce helper functions to enable/disable MDIO interface clock. This
change serves a preparatory patch for the coming feature to dynamically
control the management bus clock.
Signed-off-by: Radhey Shyam Pandey <radhey.shyam.pandey@xilinx.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Allen Pais says:
====================
net: convert tasklets to use new tasklet_setup API
Commit 12cc923f1c ("tasklet: Introduce new initialization API")'
introduced a new tasklet initialization API. This series converts
all the net/* drivers to use the new tasklet_setup() API
The following series is based on net-next (9faebeb2d)
v3:
introduce qdisc_from_priv, suggested by Eric Dumazet.
v2:
get rid of QDISC_ALIGN()
v1:
fix kerneldoc
====================
Link: https://lore.kernel.org/r/20201103091823.586717-1-allen.lkml@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Acked-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Acked-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Acked-by: Stefan Schmidt <stefan@datenfreihafen.org>
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Reviewed-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull mtd fixes from Miquel Raynal.
* 'mtd/fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux:
mtd: rawnand: stm32_fmc2: fix broken ECC
mtd: spi-nor: Fix address width on flash chips > 16MB
mtd: spi-nor: Don't copy self-pointing struct around
mtd: rawnand: ifc: Move the ECC engine initialization to the right place
mtd: rawnand: mxc: Move the ECC engine initialization to the right place
Pull spi fix from Mark Brown:
"This is an additional fix on top of 5e31ba0c05 ('spi: bcm2835: fix
gpio cs level inversion') - when sending my prior pull request I had
misremembred the status of that patch, apologies for the noise here"
* tag 'spi-fix-v5.10-rc2-2' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
spi: bcm2835: remove use of uninitialized gpio flags variable
Pull sound fixes from Takashi Iwai:
"Quite a bunch of small fixes that have been gathered since the last
pull, including changes like below:
- HD-audio runtime PM fixes and refactoring
- HD-audio and USB-audio quirks
- SOF warning fix
- Various ASoC device-specific fixes for Intel, Qualcomm, etc"
* tag 'sound-5.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (26 commits)
ALSA: usb-audio: Add implicit feedback quirk for Qu-16
ASoC: mchp-spdiftx: Do not set Validity bit(s)
ALSA: usb-audio: Add implicit feedback quirk for MODX
ALSA: usb-audio: add usb vendor id as DSD-capable for Khadas devices
ALSA: hda/realtek - Enable headphone for ASUS TM420
ALSA: hda: prevent undefined shift in snd_hdac_ext_bus_get_link()
ASoC: qcom: lpass-cpu: Fix clock disable failure
ASoC: qcom: lpass-sc7180: Fix MI2S bitwidth field bit positions
ASoC: codecs: wcd9335: Set digital gain range correctly
ASoC: codecs: wcd934x: Set digital gain range correctly
ALSA: hda: Reinstate runtime_allow() for all hda controllers
ALSA: hda: Separate runtime and system suspend
ALSA: hda: Refactor codec PM to use direct-complete optimization
ALSA: hda/realtek - Fixed HP headset Mic can't be detected
ALSA: usb-audio: Add implicit feedback quirk for Zoom UAC-2
ALSA: make snd_kcontrol_new name a normal string
ALSA: fix kernel-doc markups
ASoC: SOF: loader: handle all SOF_IPC_EXT types
ASoC: cs42l51: manage mclk shutdown delay
ASoC: qcom: sdm845: set driver name correctly
...
Pull drm fixes from Dave Airlie:
"It's Friday here so that means another installment of drm fixes to
distract you from the counting process.
Changes all over the place, the amdgpu changes contain support for a
new GPU that is close to current one already in the tree (Green
Sardine) so it shouldn't have much side effects.
Otherwise imx has a few cleanup patches and fixes, amdgpu and i915
have around the usual smattering of fixes, fonts got constified, and
vc4/panfrost has some minor fixes. All in all a fairly regular rc3.
We have an outstanding nouveau regression, but the author is looking
into the fix, so should be here next week.
I now return you to counting.
fonts:
- constify font structures.
MAINTAINERS:
- Fix path for amdgpu power management
amdgpu:
- Add support for more navi1x SKUs
- Fix for suspend on CI dGPUs
- VCN DPG fix for Picasso
- Sienna Cichlid fixes
- Polaris DPM fix
- Add support for Green Sardine
amdkfd:
- Fix an allocation failure check
i915:
- Fix set domain's cache coherency
- Fixes around breadcrumbs
- Fix encoder lookup during PSR atomic
- Hold onto an explicit ref to i915_vma_work.pinned
- gvt: HWSP reset handling fix
- gvt: flush workaround
- gvt: vGPU context pin/unpin
- gvt: mmio cmd access fix for bxt/apl
imx:
- drop unused functions and callbacks
- reuse imx_drm_encoder_parse_of
- spinlock rework
- memory leak fix
- minor cleanups
vc4:
- resource cleanup fix
panfrost:
- madvise/shrinker fix"
* tag 'drm-fixes-2020-11-06-1' of git://anongit.freedesktop.org/drm/drm: (55 commits)
drm/amdgpu/display: remove DRM_AMD_DC_GREEN_SARDINE
drm/amd/display: Add green_sardine support to DM
drm/amd/display: Add green_sardine support to DC
drm/amdgpu: enable vcn support for green_sardine (v2)
drm/amdgpu: enable green_sardine_asd.bin loading (v2)
drm/amdgpu/sdma: add sdma engine support for green_sardine (v2)
drm/amdgpu: add gfx support for green_sardine (v2)
drm/amdgpu: add soc15 common ip block support for green_sardine (v3)
drm/amdgpu: add green_sardine support for gpu_info and ip block setting (v2)
drm/amdgpu: add Green_Sardine APU flag
drm/amdgpu: resolved ASD loading issue on sienna
amdkfd: Check kvmalloc return before memcpy
drm/amdgpu: update golden setting for sienna_cichlid
amd/amdgpu: Disable VCN DPG mode for Picasso
drm/amdgpu/swsmu: remove duplicate call to smu_set_default_dpm_table
drm/i915: Hold onto an explicit ref to i915_vma_work.pinned
drm/i915/gt: Flush xcs before tgl breadcrumbs
drm/i915/gt: Expose more parameters for emitting writes into the ring
drm/i915: Fix encoder lookup during PSR atomic check
drm/i915/gt: Use the local HWSP offset during submission
...
Pull tpm fixes from Jarkko Sakkinen:
"Two critical tpm driver bug fixes"
* tag 'tpmdd-next-v5.10-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/jarkko/linux-tpmdd:
tpm: efi: Don't create binary_bios_measurements file for an empty log
tpm_tis: Disable interrupts on ThinkPad T490s
Pull iommu fixes from Joerg Roedel:
- Fix a NULL-ptr dereference in the Intel VT-d driver
- Two fixes for Intel SVM support
- Increase IRQ remapping table size in the AMD IOMMU driver. The old
number of 128 turned out to be too low for some recent devices.
- Fix a mask check in generic IOMMU code
* tag 'iommu-fixes-v5.10-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
iommu: Fix a check in iommu_check_bind_data()
iommu/vt-d: Fix a bug for PDP check in prq_event_thread
iommu/vt-d: Fix sid not set issue in intel_svm_bind_gpasid()
iommu/vt-d: Fix kernel NULL pointer dereference in find_domain()
iommu/amd: Increase interrupt remapping table limit to 512 entries
Pull VFIO fixes from Alex Williamson:
- Remove code by using existing helper (Zenghui Yu)
- fsl-mc copy-user return and underflow fixes (Dan Carpenter)
- fsl-mc static function declaration (Diana Craciun)
- Fix ioeventfd sleeping under spinlock (Alex Williamson)
- Fix pm reference count leak in vfio-platform (Zhang Qilong)
- Allow opening IGD device w/o OpRegion support (Fred Gao)
* tag 'vfio-v5.10-rc3' of git://github.com/awilliam/linux-vfio:
vfio/pci: Bypass IGD init in case of -ENODEV
vfio: platform: fix reference leak in vfio_platform_open
vfio/pci: Implement ioeventfd thread handler for contended memory lock
vfio/fsl-mc: Make vfio_fsl_mc_irqs_allocate static
vfio/fsl-mc: prevent underflow in vfio_fsl_mc_mmap()
vfio/fsl-mc: return -EFAULT if copy_to_user() fails
vfio/type1: Use the new helper to find vfio_group
Pull arm64 fixes from Will Deacon:
"Here's the weekly batch of fixes for arm64. Not an awful lot here, but
there are still a few unresolved issues relating to CPU hotplug, RCU
and IRQ tracing that I hope to queue fixes for next week.
Summary:
- Fix early use of kprobes
- Fix kernel placement in kexec_file_load()
- Bump maximum number of NUMA nodes"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: kexec_file: try more regions if loading segments fails
arm64: kprobes: Use BRK instead of single-step when executing instructions out-of-line
arm64: NUMA: Kconfig: Increase NODES_SHIFT to 4
Pull ARC fixes from Vineet Gupta:
- Unbork HSDKv1 platform (won't boot) due to memory map issue
- Prevent stack unwinder from infinite looping
* tag 'arc-5.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
ARC: [plat-hsdk] Remap CCMs super early in asm boot trampoline
ARC: stack unwinding: avoid indefinite looping
Pull s390 fixes from Heiko Carstens:
- fix reference counting for ap devices
- fix paes selftest
- fix pmd_deref()/pud_deref() so they can also handle large pages
- remove unused vdso file and defines
- update defconfigs
- call rcu_cpu_starting() early in smp init code to avoid lockdep
warnings
- fix hotplug of PCI function missing bus
* tag 's390-5.10-3' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/pci: fix hot-plug of PCI function missing bus
s390/smp: move rcu_cpu_starting() earlier
s390/pkey: fix paes selftest failure with paes and pkey static build
s390: update defconfigs
s390/vdso: remove unused constants
s390/vdso: remove empty unused file
s390/mm: make pmd/pud_deref() large page aware
s390/ap: fix ap devices reference counting
Pull networking fixes from Jakub Kicinski:
"Networking fixes for 5.10-rc3, including fixes from wireless, can, and
netfilter subtrees.
Current merge window - bugs in new features:
- can: isotp: isotp_rcv_cf(): enable RX timeout handling in
listen-only mode
Previous releases - regressions:
- mac80211:
- don't require VHT elements for HE on 2.4 GHz
- fix regression where EAPOL frames were sent in plaintext
- netfilter:
- ipset: Update byte and packet counters regardless of whether
they match
- ip_tunnel: fix over-mtu packet send by allowing fragmenting even if
inner packet has IP_DF (don't fragment) set in its header (when
TUNNEL_DONT_FRAGMENT flag is not set on the tunnel dev)
- net: fec: fix MDIO probing for some FEC hardware blocks
- ip6_tunnel: set inner ipproto before ip6_tnl_encap to un-break gso
support
- sctp: Fix COMM_LOST/CANT_STR_ASSOC err reporting on big-endian
platforms, sparse-related fix used the wrong integer size
Previous releases - always broken:
- netfilter: use actual socket sk rather than skb sk when routing
harder
- r8169: work around short packet hw bug on RTL8125 by padding frames
- net: ethernet: ti: cpsw: disable PTPv1 hw timestamping
advertisement, the hardware does not support it
- chelsio/chtls: fix always leaking ctrl_skb and another leak caused
by a race condition
- fix drivers incorrectly writing into skbs on TX:
- cadence: force nonlinear buffers to be cloned
- gianfar: Account for Tx PTP timestamp in the skb headroom
- gianfar: Replace skb_realloc_headroom with skb_cow_head for PTP
- can: flexcan:
- remove FLEXCAN_QUIRK_DISABLE_MECR quirk for LS1021A
- add ECC initialization for VF610 and LX2160A
- flexcan_remove(): disable wakeup completely
- can: fix packet echo functionality:
- peak_canfd: fix echo management when loopback is on
- make sure skbs are not freed in IRQ context in case they need to
be dropped
- always clone the skbs to make sure they have a reference on the
socket, and prevent it from disappearing
- fix real payload length return value for RTR frames
- can: j1939: return failure on bind if netdev is down, rather than
waiting indefinitely
Misc:
- IPv6: reply ICMP error if the first fragment don't include all
headers to improve compliance with RFC 8200"
* tag 'net-5.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (66 commits)
ionic: check port ptr before use
r8169: work around short packet hw bug on RTL8125
net: openvswitch: silence suspicious RCU usage warning
chelsio/chtls: fix always leaking ctrl_skb
chelsio/chtls: fix memory leaks caused by a race
can: flexcan: flexcan_remove(): disable wakeup completely
can: flexcan: add ECC initialization for VF610
can: flexcan: add ECC initialization for LX2160A
can: flexcan: remove FLEXCAN_QUIRK_DISABLE_MECR quirk for LS1021A
can: mcp251xfd: remove unneeded break
can: mcp251xfd: mcp251xfd_regmap_nocrc_read(): fix semicolon.cocci warnings
can: mcp251xfd: mcp251xfd_regmap_crc_read(): increase severity of CRC read error messages
can: peak_canfd: pucan_handle_can_rx(): fix echo management when loopback is on
can: peak_usb: peak_usb_get_ts_time(): fix timestamp wrapping
can: peak_usb: add range checking in decode operations
can: xilinx_can: handle failure cases of pm_runtime_get_sync
can: ti_hecc: ti_hecc_probe(): add missed clk_disable_unprepare() in error path
can: isotp: padlen(): make const array static, makes object smaller
can: isotp: isotp_rcv_cf(): enable RX timeout handling in listen-only mode
can: isotp: Explain PDU in CAN_ISOTP help text
...
Ido Schimmel says:
====================
nexthop: Add support for nexthop objects offload
This patch set adds support for nexthop objects offload with a dummy
implementation over netdevsim. mlxsw support will be added later.
The general idea is very similar to route offload in that notifications
are sent whenever nexthop objects are changed. A listener can veto the
change and the error will be communicated to user space with extack.
To keep listeners as simple as possible, they not only receive
notifications for the nexthop object that is changed, but also for all
the other objects affected by this change. For example, when a single
nexthop is replaced, a replace notification is sent for the single
nexthop, but also for all the nexthop groups this nexthop is member in.
This relieves listeners from the need to track such dependencies.
To simplify things further for listeners, the notification info does not
contain the raw nexthop data structures (e.g., 'struct nexthop'), but
less complex data structures into which the raw data structures are
parsed into.
Tested with a new selftest over netdevsim and with fib_nexthops.sh:
Tests passed: 164
Tests failed: 0
Patch set overview:
Patches #1-#4 introduce the aforementioned data structures and convert
existing listeners (i.e., the VXLAN driver) to use them.
Patches #5-#6 add a new RTNH_F_TRAP flag and the ability to set it and
RTNH_F_OFFLOAD on nexthops. This flag is used by netdevsim for testing
purposes and will also be used by mlxsw. These flags are consistent with
the existing RTM_F_OFFLOAD and RTM_F_TRAP flags.
Patches #7-#14 gradually add the new nexthop notifications.
Patches #15-#18 add a dummy implementation for nexthop offload over
netdevsim and a selftest to exercise both good and bad flows.
Changes since RFC [1]:
Patch #1: s/is_encap/has_encap/
Patch #3: Add a blank line in __nh_notifier_single_info_init()
Patch #5: Reword commit message
Patch #6: s/nexthop_hw_flags_set/nexthop_set_hw_flags/
Patch #7: Reword commit message
Patch #11: Allocate extack on the stack
Follow-up patch sets:
selftests: forwarding: Add nexthop objects tests
mlxsw: Preparations for nexthop objects support - part 1/2
mlxsw: Preparations for nexthop objects support - part 2/2
mlxsw: Add support for nexthop objects
mlxsw: Add support for blackhole nexthops
mlxsw: Update adjacency index more efficiently
[1] https://lore.kernel.org/netdev/20200908091037.2709823-1-idosch@idosch.org/
====================
Link: https://lore.kernel.org/r/20201104133040.1125369-1-idosch@idosch.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Test various aspects of the nexthop offload API on top of the netdevsim
implementation. Both good and bad flows are tested.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Previous patches added the ability to program nexthop objects.
Therefore, no longer forbid the programming of routes that point to such
objects.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Implement dummy nexthop "offload" in the driver by storing currently
"programmed" nexthops in a hash table. Each nexthop in the hash table is
marked with "trap" indication and increments the nexthops resource
occupancy.
This will later allow us to test the nexthop offload API on top of
netdevsim.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The Spectrum ASIC has a dedicated table where nexthops (i.e., adjacency
entries) are populated. The size of this table can be controlled via
devlink-resource.
Add such a resource to netdevsim so that its occupancy will reflect the
number of nexthop objects currently programmed to the device.
By limiting the size of the resource, error paths could be exercised and
tested.
Example output:
# devlink resource show netdevsim/netdevsim10
netdevsim/netdevsim10:
name IPv4 size unlimited unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
resources:
name fib size unlimited occ 4 unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
name fib-rules size unlimited occ 3 unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
name IPv6 size unlimited unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
resources:
name fib size unlimited occ 1 unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
name fib-rules size unlimited occ 2 unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
name nexthops size unlimited occ 0 unit entry size_min 0 size_max unlimited size_gran 1 dpipe_tables none
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Remove in-kernel route notifications when the configuration of their
nexthop changes.
These notifications are unnecessary because the route still uses the
same nexthop ID. A separate notification for the nexthop change itself
is now sent in the nexthop notification chain.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When registering a new notifier to the nexthop notification chain,
replay all the existing nexthops to the new notifier so that it will
have a complete picture of the available nexthops.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This will be used by the next patch which extends the function to replay
all the existing nexthops to the notifier block being registered.
Device drivers will be able to pass extack to the function since it is
passed to them upon reload from devlink.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When a single nexthop is deleted, the configuration of all the groups
using the nexthop is effectively modified. In this case, emit a
notification in the nexthop notification chain for each modified group
so that listeners would not need to keep track of which nexthops are
member in which groups.
In the rare cases where the notification fails, emit an error to the
kernel log. This is done by allocating extack on the stack and printing
the error logged by the listener that rejected the notification.
Changes since RFC:
* Allocate extack on the stack
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When a single nexthop is replaced, the configuration of all the groups
using the nexthop is effectively modified. In this case, emit a
notification in the nexthop notification chain for each modified group
so that listeners would not need to keep track of which nexthops are
member in which groups.
The notification can only be emitted after the new configuration (i.e.,
'struct nh_info') is pointed at by the old shell (i.e., 'struct
nexthop'). Before that the configuration of the nexthop groups is still
the same as before the replacement.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The notification is emitted after all the validation checks were
performed, but before the new configuration (i.e., 'struct nh_info') is
pointed at by the old shell (i.e., 'struct nexthop'). This prevents the
need to perform rollback in case the notification is vetoed.
The next patch will also emit a replace notification for all the nexthop
groups in which the nexthop is used.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
