M= (or KBUILD_EXTMOD) generally expects a directory path without any
trailing slashes, like M=a/b/c.
If you add a trailing slash, like M=a/b/c/, you will get ugly build
logs (two slashes in a series), but it still works fine as long as it
is consistent between 'make modules' and 'make modules_install'.
The following commands correctly build and install the modules.
$ make M=a/b/c/ modules
$ sudo make M=a/b/c/ modules_install
Since commit ccae4cfa7b ("kbuild: refactor scripts/Makefile.modinst"),
a problem happens if you add a trailing slash only for modules_install.
$ make M=a/b/c modules
$ sudo make M=a/b/c/ modules_install
No module is installed in this case, Johannes Berg reported. [1]
Trim any trailing slashes from $(KBUILD_EXTMOD).
I used the 'dirname' command to remove all the trailing slashes in
case someone adds more slashes like M=a/b/c/////. The Make's built-in
function, $(dir ...) cannot take care of such a case.
[1]: https://lore.kernel.org/lkml/10cc8522b27a051e6a9c3e158a4c4b6414fd04a0.camel@sipsolutions.net/
Fixes: ccae4cfa7b ("kbuild: refactor scripts/Makefile.modinst")
Reported-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Second set of changes for Tegra SoC memory controller drivers,
containing patchset from Thierry Reding:
"The goal here is to avoid early identity mappings altogether and instead
postpone the need for the identity mappings to when devices are attached
to the SMMU. This works by making the SMMU driver coordinate with the
memory controller driver on when to start enforcing SMMU translations.
This makes Tegra behave in a more standard way and pushes the code to
deal with the Tegra-specific programming into the NVIDIA SMMU
implementation."
This pulls a dependency from Will Deacon (ARM SMMU driver) and contains
further ARM SMMU driver patches to resolve complex dependencies between
different patchsets. The pull from Will contains only one patch
("Implement ->probe_finalize()"). Further work in Will's tree might
depend on this patch, therefore patch was applied there.
On the other hand, this ("Implement ->probe_finalize()") patch is also a
dependency for ARM SMMU driver changes for Tegra. These changes,
bringing seamless transition from the firmware framebuffer to the OS
framebuffer, depend on earlier Tegra memory controller driver patches.
-----BEGIN PGP SIGNATURE-----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=qHnq
-----END PGP SIGNATURE-----
gpgsig -----BEGIN PGP SIGNATURE-----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=pB39
-----END PGP SIGNATURE-----
Merge tag 'memory-controller-drv-tegra-5.14-2' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux-mem-ctrl into arm/drivers
Memory controller drivers for v5.14 - Tegra SoC, part two
Second set of changes for Tegra SoC memory controller drivers,
containing patchset from Thierry Reding:
"The goal here is to avoid early identity mappings altogether and instead
postpone the need for the identity mappings to when devices are attached
to the SMMU. This works by making the SMMU driver coordinate with the
memory controller driver on when to start enforcing SMMU translations.
This makes Tegra behave in a more standard way and pushes the code to
deal with the Tegra-specific programming into the NVIDIA SMMU
implementation."
This pulls a dependency from Will Deacon (ARM SMMU driver) and contains
further ARM SMMU driver patches to resolve complex dependencies between
different patchsets. The pull from Will contains only one patch
("Implement ->probe_finalize()"). Further work in Will's tree might
depend on this patch, therefore patch was applied there.
On the other hand, this ("Implement ->probe_finalize()") patch is also a
dependency for ARM SMMU driver changes for Tegra. These changes,
bringing seamless transition from the firmware framebuffer to the OS
framebuffer, depend on earlier Tegra memory controller driver patches.
* tag 'memory-controller-drv-tegra-5.14-2' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux-mem-ctrl: (37 commits)
iommu/arm-smmu: Use Tegra implementation on Tegra186
iommu/arm-smmu: tegra: Implement SID override programming
iommu/arm-smmu: tegra: Detect number of instances at runtime
dt-bindings: arm-smmu: Add Tegra186 compatible string
memory: tegra: Delete dead debugfs checking code
iommu/arm-smmu: Implement ->probe_finalize()
memory: tegra: Implement SID override programming
memory: tegra: Split Tegra194 data into separate file
memory: tegra: Add memory client IDs to tables
memory: tegra: Unify drivers
memory: tegra: Only initialize reset controller if available
memory: tegra: Make IRQ support opitonal
memory: tegra: Parameterize interrupt handler
memory: tegra: Extract setup code into callback
memory: tegra: Make per-SoC setup more generic
memory: tegra: Push suspend/resume into SoC drivers
memory: tegra: Introduce struct tegra_mc_ops
memory: tegra: Unify struct tegra_mc across SoC generations
memory: tegra: Consolidate register fields
memory: tegra30-emc: Use devm_tegra_core_dev_init_opp_table()
...
Link: https://lore.kernel.org/r/20210614195200.21657-1-krzysztof.kozlowski@canonical.com
Signed-off-by: Olof Johansson <olof@lixom.net>
- Set the fifo-size of uart_A to 128 bytes like the ARM64 counterpart
-----BEGIN PGP SIGNATURE-----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=r+Ju
-----END PGP SIGNATURE-----
gpgsig -----BEGIN PGP SIGNATURE-----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=yNYd
-----END PGP SIGNATURE-----
Merge tag 'amlogic-arm-dt-for-v5.14' of https://git.kernel.org/pub/scm/linux/kernel/git/amlogic/linux into arm/dt
Amlogic ARM DT changes for v5.14:
- Set the fifo-size of uart_A to 128 bytes like the ARM64 counterpart
* tag 'amlogic-arm-dt-for-v5.14' of https://git.kernel.org/pub/scm/linux/kernel/git/amlogic/linux:
ARM: dts: meson: Set the fifo-size of uart_A to 128 bytes
Link: https://lore.kernel.org/r/2f19a5a1-3ed9-dd11-02f1-7d535cb1c25c@baylibre.com
Signed-off-by: Olof Johansson <olof@lixom.net>
Due to a limitation for USB DFU boot mode, SPL load address has to be less
than or equal to 0x70001000. So, load address of SPL and TF-A have been
moved to 0x70000000 and 0x701c0000 respectively, in U-Boot version 2021.10.
Therefore, update TF-A's location in the device tree node.
Signed-off-by: Aswath Govindraju <a-govindraju@ti.com>
Reviewed-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Link: https://lore.kernel.org/r/20210616171224.24635-4-a-govindraju@ti.com
The final 128KB in SRAM is reserved by default for DMSC-lite code and
secure proxy communication buffer. The memory region used for DMSC-lite
code can be optionally freed up by secure firmware API[1]. However, the
buffer for secure proxy communication is not configurable. This default
hardware configuration is unique for AM64.
Therefore, indicate the area reserved for DMSC-lite code and secure proxy
communication buffer in the oc_sram device tree node.
[1] - http://downloads.ti.com/tisci/esd/latest/6_topic_user_guides/security_handover.html#triggering-security-handover
Signed-off-by: Aswath Govindraju <a-govindraju@ti.com>
Reviewed-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Link: https://lore.kernel.org/r/20210616171224.24635-3-a-govindraju@ti.com
Now that the port_groups data is being destroyed and managed by the core
code this restriction is no longer needed. All the ib_port_attrs are
compatible with the core's sysfs lifecycle.
When the main device is destroyed and moved to another namespace the
driver's port sysfs can be created/destroyed as well due to it now being a
simple attribute list.
Link: https://lore.kernel.org/r/afd8b676eace2821692d44489ff71856277c48d1.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
init_port was only being used to register sysfs attributes against the
port kobject. Now that all users are creating static attribute_group's we
can simply set the attribute_group list in the ops and the core code can
just handle it directly.
This makes all the sysfs management quite straightforward and prevents any
driver from abusing the naked port kobject in future because no driver
code can access it.
Link: https://lore.kernel.org/r/114f68f3d921460eafe14cea5a80ca65d81729c3.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
hfi1 should not be creating a mess of kobjects to attach to the port
kobject - this is all attributes. The proper API is to create an
attribute_group list and create it against the port's kobject.
Link: https://lore.kernel.org/r/cbe0ccb6175dd22274359b6ad803a37435a70e91.1623427137.git.leonro@nvidia.com
Tested-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
qib should not be creating a mess of kobjects to attach to the port
kobject - this is all attributes. The proper API is to create an
attribute_group list and create it against the port's kobject.
Link: https://lore.kernel.org/r/911e0031e1ed495b0006e8a6efec7b67a702cd5e.1623427137.git.leonro@nvidia.com
Tested-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This code is trying to attach a list of counters grouped into 4 groups to
the ib_port sysfs. Instead of creating a bunch of kobjects simply express
everything naturally as an ib_port_attribute and add a single
attribute_groups list.
Remove all the naked kobject manipulations.
Link: https://lore.kernel.org/r/0d5a7241ee0fe66622de04fcbaafaf6a791d5c7c.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Other things outside the core code are creating attributes against the
port. This patch exposes the basic machinery to do this.
The ib_port_attribute type allows creating groups of attributes attatched
to the port and comes with the usual machinery to do this.
Link: https://lore.kernel.org/r/5c4aeae57f6fa7c59a1d6d1c5506069516ae9bbf.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This call does nothing because the ib_port kobj is nested under a struct
device kobject and the dev_uevent_filter() function of the struct device
blocks uevents for any children kobj's that are not also struct devices.
A uevent for the struct device will be triggered after
ib_setup_port_attrs() returns which causes udev to pick up all the deep
"attributes" which are implemented as kobjects nested under a struct
device and assign them to the udev object for the struct device:
$ udevadm info -a /sys/class/infiniband/ibp0s9
ATTR{ports/1/counters/excessive_buffer_overrun_errors}=="0"
Link: https://lore.kernel.org/r/49231c92c7d4c60686de18f7e20932d0c82160ee.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Instead of calling device_add_groups() add the group to the existing
groups array which is managed through device_add().
This requires setting up the hw_counters before device_add(), so it gets
split up from the already split port sysfs flow.
Move all the memory freeing to the release function.
Link: https://lore.kernel.org/r/666250d937b64f6fdf45da9e2dc0b6e5e4f7abd8.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Use the same technique as gid_attrs now uses to manage the port
sysfs. Bundle everything into three allocations and use a single
sysfs_create_groups() to build everything in one shot.
All the memory is always freed in the kobj release function, removing most
of the error unwinding.
The gid_attr technique and the hw_counters are very similar, merge the two
together and combine the sysfs_create_group() call for hw_counters with
the single sysfs group setup.
Link: https://lore.kernel.org/r/b688f3340694c59f7b44b1bde40e25559ef43cf3.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Instead of having an whole bunch of different allocations to create the
gid_attr kobjects reduce it to three, one for the kobj struct plus the
attributes, and one for the attribute list for each of the two
groups.
Move the freeing of all allocations to the release function.
Reorder the operations so all the allocations happen first then the
kobject & sysfs operations are last.
This removes the majority of the complicated error unwind since the
release function will always undo all the memory allocations. Freeing the
memory is also much simpler since there is a lot less of it.
Consolidate creating the "group of array indexes" pattern into one helper
function. Ensure kobject_del is used.
Link: https://lore.kernel.org/r/f4149d379db7178d37d11d75e3026bf550f818a1.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The gid_attrs directory is a dedicated kobj nested under the port,
construct/destruct it with its own pair of functions for
understandability. This is much more readable than having it weirdly
inlined out of order into the add_port() function.
Link: https://lore.kernel.org/r/1c9434111b6770a7aef0e644a88a16eee7e325b8.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This code creates a 'struct hw_stats_attribute' for each sysfs entry that
contains a naked 'struct attribute' inside.
It then proceeds to attach this same structure to a 'struct device' kobj
and a 'struct ib_port' kobj. However, this violates the typing
requirements. 'struct device' requires the attribute to be a 'struct
device_attribute' and 'struct ib_port' requires the attribute to be
'struct port_attribute'.
This happens to work because the show/store function pointers in all three
structures happen to be at the same offset and happen to be nearly the
same signature. This means when container_of() was used to go between the
wrong two types it still managed to work.
However clang CFI detection notices that the function pointers have a
slightly different signature. As with show/store this was only working
because the device and port struct layouts happened to have the kobj at
the front.
Correct this by have two independent sets of data structures for the port
and device case. The two different attributes correctly include the
port/device_attribute struct and everything from there up is kept
split. The show/store function call chains start with device/port unique
functions that invoke a common show/store function pointer.
Link: https://lore.kernel.org/r/a8b3864b4e722aed3657512af6aa47dc3c5033be.1623427137.git.leonro@nvidia.com
Reported-by: Nathan Chancellor <nathan@kernel.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Cc: Kees Cook <keescook@chromium.org>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
It is much saner to store a pointer to the kobject structure that contains
the cannonical stats pointer than to copy the stats pointers into a public
structure.
Future patches will require the sysfs pointer for other purposes.
Link: https://lore.kernel.org/r/f90551dfd296cde1cb507bbef27cca9891d19871.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This is being used to implement both the port and device global stats,
which is causing some confusion in the drivers. For instance EFA and i40iw
both seem to be misusing the device stats.
Split it into two ops so drivers that don't support one or the other can
leave the op NULL'd, making the calling code a little simpler to
understand.
Link: https://lore.kernel.org/r/1955c154197b2a159adc2dc97266ddc74afe420c.1623427137.git.leonro@nvidia.com
Tested-by: Gal Pressman <galpress@amazon.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Implement invalidate MW and cleaned up invalidate MR operations.
Added code to perform remote invalidate for send with invalidate. Added
code to perform local invalidation. Deleted some blank lines in rxe_loc.h.
Link: https://lore.kernel.org/r/20210608042552.33275-9-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Add support for bind MW work requests from user space. Since rdma/core
does not support bind mw in ib_send_wr there is no way to support bind mw
in kernel space.
Added bind_mw local operation in rxe_req.c. Added bind_mw WR operation in
rxe_opcode.c. Added bind_mw WC in rxe_comp.c. Added additional fields to
rxe_mw in rxe_verbs.h. Added rxe_do_dealloc_mw() subroutine to cleanup an
mw when rxe_dealloc_mw is called. Added code to implement bind_mw
operation in rxe_mw.c
Link: https://lore.kernel.org/r/20210608042552.33275-8-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Simplify rxe_requester() by moving the local operations to a subroutine.
Add an error return for illegal send WR opcode. Moved next_index ahead of
rxe_run_task which fixed a small bug where work completions were delayed
until after the next wqe which was not the intended behavior. Let errors
return their own WC status. Previously all errors were reported as
protection errors which was incorrect. Changed the return of errors from
rxe_do_local_ops() to err: which causes an immediate completion. Without
this an error on a last WR may get lost. Changed fill_packet() to
finish_packet() which is more accurate.
Fixes: 8700e2e7c485 ("The software RoCE driver")
Link: https://lore.kernel.org/r/20210608042552.33275-7-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Rxe has two mask bits WR_LOCAL_MASK and WR_REG_MASK with WR_REG_MASK used
to indicate any local operation and WR_LOCAL_MASK unused. This patch
replaces both of these with one mask bit WR_LOCAL_OP_MASK which is
clearer.
Link: https://lore.kernel.org/r/20210608042552.33275-6-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Add ib_alloc_mw and ib_dealloc_mw verbs APIs.
Added new file rxe_mw.c focused on MWs. Changed the 8 bit random key
generator. Added a cleanup routine for MWs. Added verbs routines to
ib_device_ops.
Link: https://lore.kernel.org/r/20210608042552.33275-5-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Currently the rxe driver has a rxe_mw struct object but nothing about
memory windows is enabled. This patch turns on memory windows and some
minor cleanup.
Set device attribute in rxe.c so max_mw = MAX_MW. Change parameters in
rxe_param.h so that MAX_MW is the same as MAX_MR. Reduce the number of
MRs and MWs to 4K from 256K. Add device capability bits for 2a and 2b
memory windows. Removed RXE_MR_TYPE_MW from the rxe_mr_type enum.
Link: https://lore.kernel.org/r/20210608042552.33275-4-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Modify rxe_add_index() and rxe_add_key() to return an error if the index
or key is aleady present in the pool. Currently they print a warning and
silently fail with bad consequences to the caller.
Link: https://lore.kernel.org/r/20210608042552.33275-3-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Currently the rdma_rxe driver attempts to protect atomic responder
resources by taking a reference to the qp which is only freed when the
resource is recycled for a new read or atomic operation. This means that
in normal circumstances there is almost always an extra qp reference once
an atomic operation has been executed which prevents cleaning up the qp
and associated pd and cqs when the qp is destroyed.
This patch removes the call to rxe_add_ref() in send_atomic_ack() and the
call to rxe_drop_ref() in free_rd_atomic_resource(). If the qp is
destroyed while a peer is retrying an atomic op it will cause the
operation to fail which is acceptable.
Link: https://lore.kernel.org/r/20210604230558.4812-1-rpearsonhpe@gmail.com
Reported-by: Zhu Yanjun <zyjzyj2000@gmail.com>
Fixes: 86af617641 ("IB/rxe: remove unnecessary skb_clone")
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reset only the index part of the mkey and keep the variant part. On
devlink reload, driver recreates mkeys, so the mkey index may change.
Trying to preserve the variant part of the mkey, driver mistakenly
merged the mkey index with current value. In case of a devlink reload,
current value of index part is dirty, so the index may be corrupted.
Fixes: 54c62e13ad ("{IB,net}/mlx5: Setup mkey variant before mr create command invocation")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Signed-off-by: Amir Tzin <amirtz@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Running devlink reload command for port in switchdev mode cause
resources to corrupt: driver can't release allocated EQ and reclaim
memory pages, because "rdma" auxiliary device had add CQs which blocks
EQ from deletion.
Erroneous sequence happens during reload-down phase, and is following:
1. detach device - suspends auxiliary devices which support it, destroys
others. During this step "eth-rep" and "rdma-rep" are destroyed,
"eth" - suspended.
2. disable SRIOV - moves device to legacy mode; as part of disablement -
rescans drivers. This step adds "rdma" auxiliary device.
3. destroy EQ table - <failure>.
Driver shouldn't create any device during unload flows. To handle that
implement MLX5_PRIV_FLAGS_DETACH flag, set it on device detach and unset
on device attach. If flag is set do no-op on drivers rescan.
Fixes: a925b5e309 ("net/mlx5: Register mlx5 devices to auxiliary virtual bus")
Signed-off-by: Dmytro Linkin <dlinkin@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Decapsulation L3 on small inner packets which are less than
64 Bytes was done incorrectly. In small packets there is an
extra padding added in L2 which should not be included in L3
length. The issue was that after decapL3 the extra L2 padding
caused an update on the L3 length.
To avoid this issue the new header is pushed to the beginning
of the packet (offset 0) which should not cause a HW reparse
and update the L3 length.
Fixes: c349b4137c ("net/mlx5: DR, Add STEv1 modify header logic")
Reviewed-by: Erez Shitrit <erezsh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
When auxiliary bus autoprobe is disabled and SF is in ACTIVE state,
on SF port deletion it transitions from ACTIVE->ALLOCATED->INVALID.
When VHCA event handler queries the state, it is already transition
to INVALID state.
In this scenario, event handler missed to delete the SF device.
Fix it by deleting the SF when SF state is INVALID.
Fixes: 90d010b863 ("net/mlx5: SF, Add auxiliary device support")
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Vu Pham <vuhuong@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
E-switch should be able to set the GUID of host PF vport.
Currently it returns an error. This results in below error
when user attempts to configure MAC address of the PF of an
external controller.
$ devlink port function set pci/0000:03:00.0/196608 \
hw_addr 00:00:00:11:22:33
mlx5_core 0000:03:00.0: mlx5_esw_set_vport_mac_locked:1876:(pid 6715):\
"Failed to set vport 0 node guid, err = -22.
RDMA_CM will not function properly for this VF."
Check for zero vport is no longer needed.
Fixes: 330077d14d ("net/mlx5: E-switch, Supporting setting devlink port function mac address")
Signed-off-by: Yuval Avnery <yuvalav@nvidia.com>
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Bodong Wang <bodong@nvidia.com>
Reviewed-by: Alaa Hleihel <alaa@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
External controller PF's MAC address is not read from the device during
vport setup. Fail to read this results in showing all zeros to user
while the factory programmed MAC is a valid value.
$ devlink port show eth1 -jp
{
"port": {
"pci/0000:03:00.0/196608": {
"type": "eth",
"netdev": "eth1",
"flavour": "pcipf",
"controller": 1,
"pfnum": 0,
"splittable": false,
"function": {
"hw_addr": "00:00:00:00:00:00"
}
}
}
}
Hence, read it when enabling a vport.
After the fix,
$ devlink port show eth1 -jp
{
"port": {
"pci/0000:03:00.0/196608": {
"type": "eth",
"netdev": "eth1",
"flavour": "pcipf",
"controller": 1,
"pfnum": 0,
"splittable": false,
"function": {
"hw_addr": "98:03:9b:a0:60:11"
}
}
}
}
Fixes: f099fde16d ("net/mlx5: E-switch, Support querying port function mac address")
Signed-off-by: Bodong Wang <bodong@nvidia.com>
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Alaa Hleihel <alaa@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
In the case of the failure to execute mlx5_core_set_hca_defaults(),
we used wrong goto label to execute error unwind flow.
Fixes: 5bef709d76 ("net/mlx5: Enable host PF HCA after eswitch is initialized")
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Attempting to boot 32-bit ARM kernels under QEMU's 3.x virt models fails
when we have more than 512M of RAM in the model as we run out of vmalloc
space for the PCI ECAM regions. This failure will be silent when running
libvirt, as the console in that situation is a PCI device.
In this configuration, the kernel maps the whole ECAM, which QEMU sets up
for 256 buses, even when maybe only seven buses are in use. Each bus uses
1M of ECAM space, and ioremap() adds an additional guard page between
allocations. The kernel vmap allocator will align these regions to 512K,
resulting in each mapping eating 1.5M of vmalloc space. This means we need
384M of vmalloc space just to map all of these, which is very wasteful of
resources.
Fix this by only mapping the ECAM for buses we are going to be using. In
my setups, this is around seven buses in most guests, which is 10.5M of
vmalloc space - way smaller than the 384M that would otherwise be required.
This also means that the kernel can boot without forcing extra RAM into
highmem with the vmalloc= argument, or decreasing the virtual RAM available
to the guest.
Suggested-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/E1lhCAV-0002yb-50@rmk-PC.armlinux.org.uk
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Downstream Port Containment (PCIe r5.0, sec. 6.2.10) disables the link upon
an error and attempts to re-enable it when instructed by the DPC driver.
A slot which is both DPC- and hotplug-capable is currently powered off by
pciehp once DPC is triggered (due to the link change) and powered back up
on successful recovery. That's undesirable, the slot should remain powered
so the hotplugged device remains bound to its driver. DPC notifies the
driver of the error and of successful recovery in pcie_do_recovery() and
the driver may then restore the device to working state.
Moreover, Sinan points out that turning off slot power by pciehp may foil
recovery by DPC: Power off/on is a cold reset concurrently to DPC's warm
reset. Sathyanarayanan reports extended delays or failure in link
retraining by DPC if pciehp brings down the slot.
Fix by detecting whether a Link Down event is caused by DPC and awaiting
recovery if so. On successful recovery, ignore both the Link Down and the
subsequent Link Up event.
Afterwards, check whether the link is down to detect surprise-removal or
another DPC event immediately after DPC recovery. Ensure that the
corresponding DLLSC event is not ignored by synthesizing it and invoking
irq_wake_thread() to trigger a re-run of pciehp_ist().
The IRQ threads of the hotplug and DPC drivers, pciehp_ist() and
dpc_handler(), race against each other. If pciehp is faster than DPC, it
will wait until DPC recovery completes.
Recovery consists of two steps: The first step (waiting for link
disablement) is recognizable by pciehp through a set DPC Trigger Status
bit. The second step (waiting for link retraining) is recognizable through
a newly introduced PCI_DPC_RECOVERING flag.
If DPC is faster than pciehp, neither of the two flags will be set and
pciehp may glean the recovery status from the new PCI_DPC_RECOVERED flag.
The flag is zero if DPC didn't occur at all, hence DLLSC events are not
ignored by default.
pciehp waits up to 4 seconds before assuming that DPC recovery failed and
bringing down the slot. This timeout is not taken from the spec (it
doesn't mandate one) but based on a report from Yicong Yang that DPC may
take a bit more than 3 seconds on HiSilicon's Kunpeng platform.
The timeout is necessary because the DPC Trigger Status bit may never
clear: On Root Ports which support RP Extensions for DPC, the DPC driver
polls the DPC RP Busy bit for up to 1 second before giving up on DPC
recovery. Without the timeout, pciehp would then wait indefinitely for DPC
to complete.
This commit draws inspiration from previous attempts to synchronize DPC
with pciehp:
By Sinan Kaya, August 2018:
https://lore.kernel.org/linux-pci/20180818065126.77912-1-okaya@kernel.org/
By Ethan Zhao, October 2020:
https://lore.kernel.org/linux-pci/20201007113158.48933-1-haifeng.zhao@intel.com/
By Kuppuswamy Sathyanarayanan, March 2021:
https://lore.kernel.org/linux-pci/59cb30f5e5ac6d65427ceaadf1012b2ba8dbf66c.1615606143.git.sathyanarayanan.kuppuswamy@linux.intel.com/
Link: https://lore.kernel.org/r/0be565d97438fe2a6d57354b3aa4e8626952a00b.1619857124.git.lukas@wunner.de
Reported-by: Sinan Kaya <okaya@kernel.org>
Reported-by: Ethan Zhao <haifeng.zhao@intel.com>
Reported-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Tested-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Tested-by: Yicong Yang <yangyicong@hisilicon.com>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Ashok Raj <ashok.raj@intel.com>
Cc: Keith Busch <kbusch@kernel.org>
Introduce OUTSIDE macro that checks whether an instruction
address is inside or outside of a block of instructions.
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
This patch brings 3 reworked/new uevent changes:
* All AP uevents caused by an ap card or queue device now carry an
additional uevent env value MODE=<accel|cca|ep11>. Here is an
example:
KERNEL[1267.301292] add /devices/ap/card0a (ap)
ACTION=add
DEVPATH=/devices/ap/card0a
SUBSYSTEM=ap
DEVTYPE=ap_card
DEV_TYPE=000D
MODALIAS=ap:t0D
MODE=ep11 <- this is new
SEQNUM=1095
This is true for bind, unbind, add, remove, and change uevents
related to ap card or ap queue devices.
* On a change of the soft online attribute on a zcrypt queue or card
device a new CHANGE uevent is sent with an env value ONLINE=<0|1>.
Example uevent:
KERNEL[613.067531] change /devices/ap/card09/09.0011 (ap)
ACTION=change
DEVPATH=/devices/ap/card09/09.0011
SUBSYSTEM=ap
ONLINE=0 <- this is new
DEVTYPE=ap_queue
DRIVER=cex4queue
MODE=cca
SEQNUM=1070
- On a change of the config state of an zcrypt card device a new
CHANGE uevent is sent with an env value CONFIG=<0|1>.
Example uevent:
KERNEL[876.258680] change /devices/ap/card09 (ap)
ACTION=change
DEVPATH=/devices/ap/card09
SUBSYSTEM=ap
CONFIG=0 <- this is new
DEVTYPE=ap_card
DRIVER=cex4card
DEV_TYPE=000D
MODALIAS=ap:t0D
MODE=cca
SEQNUM=1073
Setting a card config on/off causes the dependent queue devices to
follow the config state change and thus uevents informing about the
config state change for the queue devices are also emitted.
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Reviewed-by: Ingo Franzki <ifranzki@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
When a AP queue is switched to soft offline, all pending
requests are purged out of the pending requests list and
'received' by the upper layer like zcrypt device drivers.
This is also done for requests which are already enqueued
into the firmware queue. A request in a firmware queue
may eventually produce an response message, but there is
no waiting process any more. However, the response was
counted with the queue_counter and as this counter was
reset to 0 with the offline switch, the pending response
caused the queue_counter to get negative. The next request
increased this counter to 0 (instead of 1) which caused
the ap code to assume there is nothing to receive and so
the response for this valid request was never tried to
fetch from the firmware queue.
This all caused a queue to not work properly after a
switch offline/online and in the end processes to hang
forever when trying to send a crypto request after an
queue offline/online switch cicle.
Fixed by a) making sure the counter does not drop below 0
and b) on a successful enqueue of a message has at least
a value of 1.
Additionally a warning is emitted, when a reply can't get
assigned to a waiting process. This may be normal operation
(process had timeout or has been killed) but may give a
hint that something unexpected happened (like this odd
behavior described above).
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Cc: stable@vger.kernel.org
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Merge __calc_map_type_and_dist() and calc_map_type_and_dist_warn() into
calc_map_type_and_dist() to simplify the code a bit. This now means we add
the devfn strings to the acs_buf unconditionally even if the buffer is not
printed, but that is not a lot of overhead and keeps the code much simpler.
Link: https://lore.kernel.org/r/20210614055310.3960791-1-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Pablo says, tprot_set is only there to detect if tprot was set to
IPPROTO_IP as that evaluates to zero. Therefore, code asserting a
different value in tprot does not need to check tprot_set.
Fixes: 935b7f6430 ("netfilter: nft_exthdr: add TCP option matching")
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>