Convert the LAN78xx USB Ethernet driver to use the PHYLINK framework for
managing PHY and MAC interactions. This improves consistency with other
network drivers, simplifies pause frame handling, and enables cleaner
suspend/resume support.
Key changes:
- Replace all PHYLIB-based logic with PHYLINK equivalents:
- Replace phy_connect()/phy_disconnect() with phylink_connect_phy()
- Replace phy_start()/phy_stop() with phylink_start()/phylink_stop()
- Replace pauseparam handling with phylink_ethtool_get/set_pauseparam()
- Introduce lan78xx_phylink_setup() to configure PHYLINK
- Add phylink MAC operations:
- lan78xx_mac_config()
- lan78xx_mac_link_up()
- lan78xx_mac_link_down()
- Remove legacy link state handling:
- lan78xx_link_status_change()
- lan78xx_link_reset()
- Handle fixed-link fallback for LAN7801 using phylink_set_fixed_link()
- Replace deprecated flow control handling with phylink-managed logic
Power management:
- Switch suspend/resume paths to use phylink_suspend()/phylink_resume()
- Ensure proper use of rtnl_lock() where required
- Note: full runtime testing of power management is currently limited
due to hardware setup constraints
Note: Conversion of EEE (Energy Efficient Ethernet) handling to the
PHYLINK-managed API will be done in a follow-up patch. For now, the
legacy EEE enable logic is preserved in mac_link_up().
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Link: https://patch.msgid.link/20250618122602.3156678-2-o.rempel@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We can avoid open-coding the loop construct which counts firmware child
nodes with a specific name by using the newly added
device_get_named_child_node_count().
The gianfar driver has such open-coded loop. Replace it with the
device_get_child_node_count_named().
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Matti Vaittinen <mazziesaccount@gmail.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/3a33988fc042588cb00a0bfc5ad64e749cb0eb1f.1750248902.git.mazziesaccount@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In order to clean up of the VLAN handling, factor out the VLAN
handling into separate function fec_enet_rx_vlan().
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-11-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move __vlan_hwaccel_put_tag() into the if statement that sets
vlan_packet_rcvd = true. This change eliminates the unnecessary
vlan_packet_rcvd variable, simplifying the code and improving clarity.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-10-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In order to clean up of the VLAN handling, reduce the scope of data.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-9-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For better readability and maintainability, use the provided helper function
skb_vlan_eth_hdr() to replace manual the VLAN header calculation, and change
the type of vlan_header to struct vlan_ethhdr to take into account that the
Ethernet header plus VLAN header is returned.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-8-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There are the functions fec_enet_rx_queue() and fec_enet_tx_queue(),
one for handling the RX queue the other one handles the TX queue.
However they don't have the same signature. Align fec_enet_rx_queue()
argument order with fec_enet_tx_queue() to make code more readable.
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-7-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Replace "1 << 5" for configuring 1000 MBit/s with a defined constant to
improve code readability and maintainability.
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-6-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In da722186f6 ("net: fec: set GPR bit on suspend by DT
configuration.") the platform_device_id fec_devtype::driver_data was
converted from holding the quirks to a pointing to struct fec_devinfo.
The struct fec_devinfo holding the information for the i.MX6SX was
named fec_imx6x_info.
Rename fec_imx6x_info to fec_imx6sx_info to align with the SoC's name.
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-5-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This is a preparation patch to make addition of new includes easier
without breaking the alphabetic order.
Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-4-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In commit 4d494cdc92 ("net: fec: change data structure to support
multiqueue") the data structures were changed, so that the comment about
the sent-in-place skb doesn't apply any more. Remove it.
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-2-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
codespell has found some typos in the comments, fix them.
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Csókás, Bence <csokas.bence@prolan.hu>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://patch.msgid.link/20250618-fec-cleanups-v4-1-c16f9a1af124@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Russell King says:
====================
net: stmmac: loongson1: cleanups
A couple of patches to cleanup loongson1. First, introducing a match
data struct to allow the per-match data to be extended beyond the init
function pointer, and then adding a setup method to allow the resource
base address to be translated to the MAC index at probe time rather
than repeatedly in the setup function.
====================
Link: https://patch.msgid.link/aFKXzlno7HkG-cNh@shell.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ls1b_dwmac_syscon_init() was getting the stmmac iomem resource to detect
which GMAC block is being used. Move this to a separate setup() function
that only runs at probe time, so it can sensibly behave with an
unrecognised resource adress. Use this to set a MAC index (id) which is
then used in place of testing the base address.
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Keguang Zhang <keguang.zhang@gmail.com>
Tested-by: Keguang Zhang <keguang.zhang@gmail.com> # on LS1B & LS1C
Link: https://patch.msgid.link/E1uRqEE-004c7M-Go@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Provide a structure for match data rather than using the function
pointer as match data. This allows stronger type-checking for the
function itself, and allows extensions to the match data.
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Keguang Zhang <keguang.zhang@gmail.com>
Tested-by: Keguang Zhang <keguang.zhang@gmail.com> # on LS1B & LS1C
Link: https://patch.msgid.link/E1uRqE9-004c7G-CB@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pass the stmmac_priv structure into the pcs_set_ane() MAC method rather
than having callers dereferencing this structure for the IO address.
Tested-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org> # sa8775p-ride-r3
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/E1uRqbQ-004djP-1l@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add power-domain nodes for the power controller on RK3528.
Only PD_GPU can fully be powered down. PD_RKVDEC, PD_RKVENC, PD_VO and
PD_VPU are idle only power domains used by miscellaneous devices.
Because multiple of the miscellaneous device types currently complain
about the use of a power-domains prop, only PD_GPU is enabled.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Link: https://lore.kernel.org/r/20250518220707.669515-5-jonas@kwiboo.se
[changed to using numeric values, until the next merge-window]
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
The ArmSoM Sige5 has several USB ports: a Type-A USB 3 port (USB2 lines
going through a hub), a Type-A USB 2.0 port (also going through a hub),
a Type-C DC input port that has absolutely no USB data connection and a
Type-C port with USB3.2 Gen1x1 that's also the maskrom programming port.
Enable these ports, and set the device role to be host for the host
ports.
The data capable Type-C USB port uses a fusb302 for data role switching.
Signed-off-by: Nicolas Frattaroli <nicolas.frattaroli@collabora.com>
Link: https://lore.kernel.org/r/20250619-rk3576-sige5-usb-v5-2-9069a7e750e1@collabora.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
As reported by syzbot, hashtab_init() can be affected by abnormally
large policy loads which would cause the kernel's allocator to emit
a warning in some configurations. Since the SELinux hashtab_init()
code handles the case where the allocation fails, due to a large
request or some other reason, we can safely add the __GFP_NOWARN flag
to squelch these abnormally large allocation warnings.
Reported-by: syzbot+bc2c99c2929c3d219fb3@syzkaller.appspotmail.com
Tested-by: syzbot+bc2c99c2929c3d219fb3@syzkaller.appspotmail.com
Signed-off-by: Paul Moore <paul@paul-moore.com>
Extend the task avdcache to also cache whether the task SID is both
permissive and neveraudit, and return immediately if so in both
selinux_inode_getattr() and selinux_inode_permission().
The same approach could be applied to many of the hook functions
although the avdcache would need to be updated for more than directory
search checks in order for this optimization to be beneficial for checks
on objects other than directories.
To test, apply https://github.com/SELinuxProject/selinux/pull/473 to
your selinux userspace, build and install libsepol, and use the following
CIL policy module:
$ cat neverauditpermissive.cil
(typeneveraudit unconfined_t)
(typepermissive unconfined_t)
Without this module inserted, running the following commands:
perf record make -jN # on an already built allmodconfig tree
perf report --sort=symbol,dso
yields the following percentages (only showing __d_lookup_rcu for
reference and only showing relevant SELinux functions):
1.65% [k] __d_lookup_rcu
0.53% [k] selinux_inode_permission
0.40% [k] selinux_inode_getattr
0.15% [k] avc_lookup
0.05% [k] avc_has_perm
0.05% [k] avc_has_perm_noaudit
0.02% [k] avc_policy_seqno
0.02% [k] selinux_file_permission
0.01% [k] selinux_inode_alloc_security
0.01% [k] selinux_file_alloc_security
for a total of 1.24% for SELinux compared to 1.65% for
__d_lookup_rcu().
After running the following command to insert this module:
semodule -i neverauditpermissive.cil
and then re-running the same perf commands from above yields
the following non-zero percentages:
1.74% [k] __d_lookup_rcu
0.31% [k] selinux_inode_permission
0.03% [k] selinux_inode_getattr
0.03% [k] avc_policy_seqno
0.01% [k] avc_lookup
0.01% [k] selinux_file_permission
0.01% [k] selinux_file_open
for a total of 0.40% for SELinux compared to 1.74% for
__d_lookup_rcu().
Signed-off-by: Stephen Smalley <stephen.smalley.work@gmail.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Introduce neveraudit types i.e. types that should never trigger
audit messages. This allows the AVC to skip all audit-related
processing for such types. Note that neveraudit differs from
dontaudit not only wrt being applied for all checks with a given
source type but also in that it disables all auditing, not just
permission denials.
When a type is both a permissive type and a neveraudit type,
the security server can short-circuit the security_compute_av()
logic, allowing all permissions and not auditing any permissions.
This change just introduces the basic support but does not yet
further optimize the AVC or hook function logic when a type
is both a permissive type and a dontaudit type.
Suggested-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Stephen Smalley <stephen.smalley.work@gmail.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
List both CPU supply regulators which drive the little and big CPU
clusters, respectively, so that cpufreq can pick them up.
Without this patch the cpufreq governor attempts to raise the big CPU
frequency under high load, while its supply voltage stays at 850000 uV.
This causes system instability and, in my case, random reboots.
With this patch, supply voltages are adjusted in step with frequency
changes from 700000-737000 uV in idle to 950000 uV under full load,
and the system appears to be stable.
While at this, list all CPU supplies for completeness.
Cc: stable@vger.kernel.org
Fixes: 40f742b07a ("arm64: dts: rockchip: Add rk3576-armsom-sige5 board")
Reviewed-by: Nicolas Frattaroli <nicolas.frattaroli@collabora.com>
Tested-by: Nicolas Frattaroli <nicolas.frattaroli@collabora.com>
Signed-off-by: Alexey Charkov <alchark@gmail.com>
Link: https://lore.kernel.org/r/20250614-sige5-updates-v2-1-3bb31b02623c@gmail.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Add support for the Broadcom based WiFi/Bluetooth module (BW3752-50B1)
found in ArmSoM Sige5 boards version 1.2. This includes SDIO connected
WiFi with OOB interrupt support, as well as UART connected Bluetooth
with its respective interrupts.
PCM support for Bluetooth SCO audio is left out for now. It is connected
to SAI2 in M0 pin mode in case someone needs to enable it.
Note that v1.1 boards used a Realtek based module which is incompatible
with these DT nodes, so v1.1 would need a different overlay.
Signed-off-by: Alexey Charkov <alchark@gmail.com>
Link: https://lore.kernel.org/r/20250614-sige5-updates-v2-4-3bb31b02623c@gmail.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
ArmSoM Sige5 uses a soldered-on WiFi/BT module with WiFi on SDIO and BT
on UART. However, board v1.1 uses a Realtek based BL-M8852BS2, while
v1.2 uses a Broadcom based BW3752-50B1. They use the same pins and
controllers, but require different DT properties to enable.
Thankfully, the WiFi part at least works without explicitly listing it in
the device tree, albeit without OOB interrupt functionality.
Add required device tree nodes that do not depend on the board version so
that at least the WiFi module can appear on the SDIO bus.
WiFi OOB interrupt and Bluetooth function support are not enabled here, as
they require module specific properties.
Signed-off-by: Alexey Charkov <alchark@gmail.com>
Link: https://lore.kernel.org/r/20250614-sige5-updates-v2-3-3bb31b02623c@gmail.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
RK3576 has one more SD/MMC controller than are currently listed in its
.dtsi, with the missing one intended as an SDIO controller. Add the
missing node (tested with the onboard WiFi module on ArmSoM Sige5 v1.2)
Signed-off-by: Alexey Charkov <alchark@gmail.com>
Link: https://lore.kernel.org/r/20250614-sige5-updates-v2-2-3bb31b02623c@gmail.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
If the end result of a security_compute_sid() computation matches the
ssid or tsid, return that SID rather than looking it up again. This
avoids the problem of multiple initial SIDs that map to the same
context.
Cc: stable@vger.kernel.org
Reported-by: Guido Trentalancia <guido@trentalancia.com>
Fixes: ae254858ce ("selinux: introduce an initial SID for early boot processes")
Signed-off-by: Stephen Smalley <stephen.smalley.work@gmail.com>
Tested-by: Guido Trentalancia <guido@trentalancia.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Convert lpc1850-ccu.txt to yaml format.
Additional changes:
- remove label in examples.
- remove clock consumer in examples.
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Link: https://lore.kernel.org/r/20250602141937.942091-1-Frank.Li@nxp.com
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Fix a few issues in the panel section of the PinePhone Pro DTS:
- add the second part of the Himax HX8394 LCD panel controller
compatible
- as proposed by Diederik de Haas, reuse the mipi_out and ports
definitions from rk3399-base.dtsi instead of redefining them
- add a pinctrl for the LCD_RST signal for LCD1, derived from
LCD1_RST, which is on GPIO4_D1, as documented on pages 11
and 16 of the PinePhone Pro schematic
Signed-off-by: Olivier Benjamin <olivier.benjamin@bootlin.com>
Reviewed-by: Diederik de Haas <didi.debian@cknow.org>
Link: https://lore.kernel.org/r/20250619-dtb_fixes-v3-1-9cb02ddd8ce4@bootlin.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
When the first driver for Apple Silicon was upstreamed we accidentally
included `default ARCH_APPLE` in its Kconfig which then spread to almost
every subsequent driver. As soon as ARCH_APPLE is set to y this will
pull in many drivers as built-ins which is not what we want.
Thus, drop `default ARCH_APPLE` from Kconfig.
Signed-off-by: Sven Peter <sven@kernel.org>
Link: https://lore.kernel.org/r/20250612-apple-kconfig-defconfig-v1-3-0e6f9cb512c1@kernel.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Converts `if-else` blocks into one line code using `map_or`
for simplicity.
Signed-off-by: Onur Özkan <work@onurozkan.dev>
Link: https://lore.kernel.org/r/20250618093508.16343-1-work@onurozkan.dev
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Marks `Hertz` methods as `const` to make them available
for `const` contexts. This can be useful when defining
static/compile-time frequency parameters in drivers/subsystems.
Signed-off-by: Onur Özkan <work@onurozkan.dev>
Link: https://lore.kernel.org/r/20250618091442.29104-1-work@onurozkan.dev
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Now the new ucmd-based object allocator eases the finalize/abort routine,
apply this to all existing allocators that aren't protected by any lock.
Upgrade the for-driver vIOMMU alloctor too, and pass down to all existing
viommu_alloc op accordingly.
Note that __iommufd_object_alloc_ucmd() builds in some static tests that
cover both static_asserts in the iommufd_viommu_alloc(). Thus drop them.
Link: https://patch.msgid.link/r/107b24a3b791091bb09c92ffb0081c56c413b26d.1749882255.git.nicolinc@nvidia.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
An object allocator needs to call either iommufd_object_finalize() upon a
success or iommufd_object_abort_and_destroy() upon an error code.
To reduce duplication, store a new_obj in the struct iommufd_ucmd and call
iommufd_object_finalize/iommufd_object_abort_and_destroy() accordingly in
the main function.
Similar to iommufd_object_alloc() and __iommufd_object_alloc(), add a pair
of helpers: __iommufd_object_alloc_ucmd() and iommufd_object_alloc_ucmd().
Link: https://patch.msgid.link/r/e7206d4227844887cc8dbf0cc7b0242580fafd9d.1749882255.git.nicolinc@nvidia.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Now, all driver structures will be allocated by the core, i.e. no longer a
need of driver calling _iommufd_object_alloc. Thus, move it back.
Before:
text data bss dec hex filename
3024 180 0 3204 c84 drivers/iommu/iommufd/driver.o
9074 610 64 9748 2614 drivers/iommu/iommufd/main.o
After:
text data bss dec hex filename
2665 164 0 2829 b0d drivers/iommu/iommufd/driver.o
9410 618 64 10092 276c drivers/iommu/iommufd/main.o
Link: https://patch.msgid.link/r/79e630c7b911930cf36e3c8a775a04e66c528d65.1749882255.git.nicolinc@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
To ease the for-driver iommufd APIs, get_viommu_size and viommu_init ops
are introduced. Now, those existing vIOMMU supported drivers implemented
these two ops, replacing the viommu_alloc one. So, there is no use of it.
Remove it from the headers and the viommu core.
Link: https://patch.msgid.link/r/5b32d4499d7ed02a63e57a293c11b642d226ef8d.1749882255.git.nicolinc@nvidia.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
To ease the for-driver iommufd APIs, get_viommu_size and viommu_init ops
are introduced.
Sanitize the inputs and report the size of struct arm_vsmmu on success, in
arm_smmu_get_viommu_size().
Place the type sanity at the last, becase there will be soon an impl level
get_viommu_size op, which will require the same sanity tests prior. It can
simply insert a piece of code in front of the IOMMU_VIOMMU_TYPE_ARM_SMMUV3
sanity.
The core will ensure the viommu_type is set to the core vIOMMU object, and
pass in the same dev pointer, so arm_vsmmu_init() won't need to repeat the
same sanity tests but to simply init the arm_vsmmu struct.
Remove the arm_vsmmu_alloc, completing the replacement.
Link: https://patch.msgid.link/r/64e4b4c33acd26e1bd676e077be80e00fb63f17c.1749882255.git.nicolinc@nvidia.com
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
To ease the for-driver iommufd APIs, get_viommu_size and viommu_init ops
are introduced.
Sanitize the inputs and report the size of struct mock_viommu on success,
in mock_get_viommu_size().
The core will ensure the viommu_type is set to the core vIOMMU object, so
simply init the driver part in mock_viommu_init().
Remove the mock_viommu_alloc, completing the replacement.
Link: https://patch.msgid.link/r/993beabbb0bc9705d979a92801ea5ed5996a34eb.1749882255.git.nicolinc@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
There is no use of this parent domain. Delete the dead code.
Note that the s2_parent in struct mock_viommu will be a deadcode too. Yet,
keep it because it will be soon used by HW queue objects, i.e. no point in
adding it back and forth in such a short window. Besides, keeping it could
cover the majority of vIOMMU use cases where a driver-level structure will
be larger in size than the core structure.
Link: https://patch.msgid.link/r/0f155a7cd71034a498448fe4828fb4aaacdabf95.1749882255.git.nicolinc@nvidia.com
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
To ease the for-driver iommufd APIs, get_viommu_size and viommu_init ops
are introduced to replace the viommu_init op.
Let the new viommu_init pathway coexist with the old viommu_alloc one.
Since the viommu_alloc op and its pathway will be soon deprecated, try to
minimize the code difference between them by adding a tentative jump tag.
Note that this fails a !viommu->ops case from now on with a WARN_ON_ONCE
since a vIOMMU is expected to support an alloc_domain_nested op for now,
or some sort of a viommu op in the foreseeable future. This WARN_ON_ONCE
can be lifted, if some day there is a use case wanting !viommu->ops.
Link: https://patch.msgid.link/r/35c5fa5926be45bda82f5fc87545cd3180ad4c9c.1749882255.git.nicolinc@nvidia.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
So far, a vIOMMU object has been allocated by IOMMU driver and initialized
with the driver-level structure, before it returns to the iommufd core for
core-level structure initialization. It has been requiring iommufd core to
expose some core structure/helpers in its driver.c file, which result in a
size increase of this driver module.
Meanwhile, IOMMU drivers are now requiring more vIOMMU-base structures for
some advanced feature, such as the existing vDEVICE and a future HW_QUEUE.
Initializing a core-structure later than driver-structure gives for-driver
helpers some trouble, when they are used by IOMMU driver assuming that the
new structure (including core) are fully initialized, for example:
core: viommu = ops->viommu_alloc();
driver: // my_viommu is successfully allocated
driver: my_viommu = iommufd_viommu_alloc(...);
driver: // This may crash if it reads viommu->ictx
driver: new = iommufd_new_viommu_helper(my_viommu->core ...);
core: viommu->ictx = ucmd->ictx;
core: ...
To ease such a condition, allow the IOMMU driver to report the size of its
vIOMMU structure, let the core allocate a vIOMMU object and initialize the
core-level structure first, and then hand it over the driver to initialize
its driver-level structure.
Thus, this requires two new iommu ops, get_viommu_size and viommu_init, so
iommufd core can communicate with drivers to replace the viommu_alloc op:
core: viommu = ops->get_viommu_size();
driver: return VIOMMU_STRUCT_SIZE();
core: viommu->ictx = ucmd->ictx; // and others
core: rc = ops->viommu_init();
driver: // This is safe now as viommu->ictx is inited
driver: new = iommufd_new_viommu_helper(my_viommu->core ...);
core: ...
This also adds a VIOMMU_STRUCT_SIZE macro, for drivers to use, which would
statically sanitize the driver structure.
Link: https://patch.msgid.link/r/3ab52c5b622dad476c43b1b1f1636c8b902f1692.1749882255.git.nicolinc@nvidia.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
It's more accurate to report EOPNOTSUPP when an ioctl failed due to driver
bug, since there is nothing wrong with the user space side.
Link: https://patch.msgid.link/r/623bb6f0e8fdd7b9c5745a2f99f280163f9f1f5a.1749882255.git.nicolinc@nvidia.com
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Replace unsigned int, to make it clear. No functional changes.
Link: https://patch.msgid.link/r/208a260c100a00667d3799feaad1260745f96c6b.1749882255.git.nicolinc@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Replace unsigned int, to make it clear. No functional changes.
The viommu_alloc iommu op will be deprecated, so don't change that.
Link: https://patch.msgid.link/r/6c6ba5c0cd381594f17ae74355872d78d7a022c0.1749882255.git.nicolinc@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The core code can always get the ictx pointer via vdev->viommu->ictx, thus
drop this unused one.
Link: https://patch.msgid.link/r/6cbb65e8df433de45b6c3a4bb2c5df09faca8a7c.1749882255.git.nicolinc@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Run clang-format but exclude those not so obvious ones, which leaves us:
- Align indentations
- Add missing spaces
- Remove unnecessary spaces
- Remove unnecessary line wrappings
Link: https://patch.msgid.link/r/9132e1ab45690ab1959c66bbb51ac5536a635388.1749882255.git.nicolinc@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
