On Mon, Mar 17, 2025 at 06:57:01PM -0700, Chris Torek wrote:
> On Mon, Mar 17, 2025 at 11:38 AM Willy Tarreau wrote:
> > OK thanks, but that remains quite strange to me. How can we end up
> > here with such an unaligned stack ? At the very minimum I'd expect
> > all offsets to be multiple of 8.
>
Currently, the only user of vhost_log_write() is vhost-net. The 'len'
argument prevents logging of pages that are not tainted by the RX path.
Adjustments are needed since more drivers (i.e. vhost-scsi) begin using
vhost_log_write(). So far vhost-net RX path may only partially use pages
shared by t
In a multi-peer scenario there are a number of situations when a
specific peer needs to be looked up.
We may want to lookup a peer by:
1. its ID
2. its VPN destination IP
3. its transport IP/port couple
For each of the above, there is a specific routing table referencing all
peers for fast look u
This change introduces the netlink command needed to add, delete and
retrieve/dump known peers. Userspace is expected to use these commands
to handle known peer lifecycles.
Signed-off-by: Antonio Quartulli
---
drivers/net/ovpn/netlink.c | 678 -
driver
With this change ovpn is allowed to communicate to peers also via TCP.
Parsing of incoming messages is implemented through the strparser API.
Note that ovpn redefines sk_prot and sk_socket->ops for the TCP socket
used to communicate with the peer.
For this reason it needs to access inet6_stream_op
Byte/packet counters for in-tunnel and transport streams
are now initialized and updated as needed.
To be exported via netlink.
Signed-off-by: Antonio Quartulli
---
drivers/net/ovpn/Makefile | 1 +
drivers/net/ovpn/io.c | 12 +++-
drivers/net/ovpn/peer.c | 2 ++
drivers/net/ovpn
Packets sent over the ovpn interface are processed and transmitted to the
connected peer, if any.
Implementation is UDP only. TCP will be added by a later patch.
Note: no crypto/encapsulation exists yet. Packets are just captured and
sent.
Signed-off-by: Antonio Quartulli
---
drivers/net/Kconf
On Mon, Mar 17, 2025 at 11:38 AM Willy Tarreau wrote:
> OK thanks, but that remains quite strange to me. How can we end up
> here with such an unaligned stack ? At the very minimum I'd expect
> all offsets to be multiple of 8.
It's a peculiar feature of the version 9 SPARC architecture and runtim
Userspace may want to pass the MSG_NOSIGNAL flag to
tcp_sendmsg() in order to avoid generating a SIGPIPE.
To pass this flag down the TCP stack a new skb sending API
accepting a flags argument is introduced.
Cc: Eric Dumazet
Cc: Paolo Abeni
Signed-off-by: Antonio Quartulli
---
drivers/net/ovpn
Packets received over the socket are forwarded to the user device.
Implementation is UDP only. TCP will be added by a later patch.
Note: no decryption/decapsulation exists yet, packets are forwarded as
they arrive without much processing.
Signed-off-by: Antonio Quartulli
---
drivers/net/ovpn/i
This specific structure is used in the ovpn kernel module
to wrap and carry around a standard kernel socket.
ovpn takes ownership of passed sockets and therefore an ovpn
specific objects is attached to them for status tracking
purposes.
Initially only UDP support is introduced. TCP will come in a
When sending an skb over a socket using skb_send_sock_locked(),
it is currently not possible to specify any flag to be set in
msghdr->msg_flags.
However, we may want to pass flags the user may have specified,
like MSG_NOSIGNAL.
Extend __skb_send_sock() with a new argument 'flags' and add a
new in
This change implements encryption/decryption and
encapsulation/decapsulation of OpenVPN packets.
Support for generic crypto state is added along with
a wrapper for the AEAD crypto kernel API.
Signed-off-by: Antonio Quartulli
---
drivers/net/Kconfig| 4 +
drivers/net/ovpn/Makefile
The ovpn-cli tool can be compiled and used as selftest for the ovpn
kernel module.
[NOTE: it depends on libmedtls for decoding base64-encoded keys]
ovpn-cli implements the netlink and RTNL APIs and can thus be integrated
in any script for more automated testing.
Along with the tool, a bunch of s
Implement support for basic ethtool functionality.
Note that ovpn is a virtual device driver, therefore
various ethtool APIs are just not meaningful and thus
not implemented.
Signed-off-by: Antonio Quartulli
Reviewed-by: Andrew Lunn
---
drivers/net/ovpn/main.c | 15 +++
1 file chan
IV wrap-around is cryptographically dangerous for a number of ciphers,
therefore kill the key and inform userspace (via netlink) should the
IV space go exhausted.
Userspace has two ways of deciding when the key has to be renewed before
exhausting the IV space:
1) time based approach:
after X se
This change introduces the netlink commands needed to add, get, delete
and swap keys for a specific peer.
Userspace is expected to use these commands to create, inspect (non
sensitive data only), destroy and rotate session keys for a specific
peer.
Signed-off-by: Antonio Quartulli
---
drivers/n
OpenVPN is a userspace software existing since around 2005 that allows
users to create secure tunnels.
So far OpenVPN has implemented all operations in userspace, which
implies several back and forth between kernel and user land in order to
process packets (encapsulate/decapsulate, encrypt/decrypt
In case of UDP links, the local or remote endpoint used to communicate
with a given peer may change without a connection restart.
Add support for learning the new address in case of change.
Signed-off-by: Antonio Quartulli
---
drivers/net/ovpn/io.c | 8 ++
drivers/net/ovpn/peer.c | 213
With this change an ovpn instance will be able to stay connected to
multiple remote endpoints.
This functionality is strictly required when running ovpn on an
OpenVPN server.
Signed-off-by: Antonio Quartulli
---
drivers/net/ovpn/main.c | 64 +--
drivers/net/ovpn/ovpnpriv.h | 1
Add basic infrastructure for handling ovpn interfaces.
Tested-by: Donald Hunter
Signed-off-by: Antonio Quartulli
---
Documentation/netlink/specs/rt_link.yaml | 16 ++
drivers/net/ovpn/Makefile| 1 +
drivers/net/ovpn/io.c| 22 +++
drivers/net/ovpn/io.
An ovpn interface configured in MP mode will keep carrier always
on and let the user decide when to bring it administratively up and
down.
This way a MP node (i.e. a server) will keep its interface always
up and running, even when no peer is connected.
Signed-off-by: Antonio Quartulli
---
drive
This commit introduces basic netlink support with family
registration/unregistration functionalities and stub pre/post-doit.
More importantly it introduces the YAML uAPI description along
with its auto-generated files:
- include/uapi/linux/ovpn.h
- drivers/net/ovpn/netlink-gen.c
- drivers/net/ovpn
Notable changes since v23:
* dropped call to netif_tx_start/stop_all_queues()
* dropped NETIF_F_HW_CSUM and NETIF_F_RXCSUM dev flags
* dropped conditional call to skb_checksum_help() due to the point above
* added call to dst_cache_reset() in nl_peer_modify()
* dropped obsolete comment in ovpn_peer
On Tue, Mar 18, 2025 at 7:51 AM Dongli Zhang wrote:
>
> Currently, the only user of vhost_log_write() is vhost-net. The 'len'
> argument prevents logging of pages that are not tainted by the RX path.
>
> Adjustments are needed since more drivers (i.e. vhost-scsi) begin using
> vhost_log_write(). S
On Tue, Mar 18, 2025 at 7:52 AM Dongli Zhang wrote:
>
> Although the support of VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 was
> signaled by the commit 664ed90e621c ("vhost/scsi: Set
> VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 feature bits"),
> vhost_scsi_send_bad_target() still assumes the response
On Tue, Mar 18, 2025 at 7:52 AM Dongli Zhang wrote:
>
> Although the support of VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 was
> signaled by the commit 664ed90e621c ("vhost/scsi: Set
> VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 feature bits"),
> vhost_scsi_send_bad_target() still assumes the response
Although the support of VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 was
signaled by the commit 664ed90e621c ("vhost/scsi: Set
VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 feature bits"),
vhost_scsi_send_bad_target() still assumes the response in a single
descriptor.
In addition, although vhost_scsi_send_
On Tue, Mar 18, 2025 at 7:51 AM Dongli Zhang wrote:
>
> The vhost-scsi completion path may access vq->log_base when vq->log_used is
> already set to false.
>
> vhost-thread QEMU-thread
>
> vhost_scsi_complete_cmd_work()
> -> vhost_add_used()
>-> vhost_add_used_n()
>
Fix a typo in an email address.
Cc: sta...@vger.kernel.org
Reported-by: Konstantin Ryabitsev
Closes:
https://lore.kernel.org/all/20240925-rational-succinct-vulture-cca9fb@lemur/T/
Signed-off-by: Alexey Makhalov
---
MAINTAINERS | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff -
Hi Mike,
On 3/17/25 4:55 PM, Dongli Zhang wrote:
> The vhost-scsi I/O queue uses vhost_scsi_cmd. Allocate the log buffer
> during vhost_scsi_cmd allocation or when VHOST_F_LOG_ALL is set. Free the
> log buffer when vhost_scsi_cmd is reclaimed or when VHOST_F_LOG_ALL is
> removed.
>
> Fail vhost_s
Simplify of_dma_set_restricted_buffer() by using of_property_present()
and of_for_each_phandle() iterator.
Signed-off-by: Rob Herring (Arm)
---
drivers/of/device.c | 34 +-
1 file changed, 13 insertions(+), 21 deletions(-)
diff --git a/drivers/of/device.c b/drive
Log write descriptors for the event queue, leveraging vhost_get_vq_desc()
to retrieve the array of write descriptors to obtain the log buffer.
There is only one path for event queue.
Suggested-by: Joao Martins
Signed-off-by: Dongli Zhang
---
drivers/vhost/scsi.c | 16 +++-
1 file c
The vhost-scsi I/O queue uses vhost_scsi_cmd. Allocate the log buffer
during vhost_scsi_cmd allocation or when VHOST_F_LOG_ALL is set. Free the
log buffer when vhost_scsi_cmd is reclaimed or when VHOST_F_LOG_ALL is
removed.
Fail vhost_scsi_set_endpoint or vhost_scsi_set_features() on allocation
fa
Since long time ago, the only user of vq->log is vhost-net. The concern is
to add support for more devices (i.e. vhost-scsi or vsock) may reveals
unknown issue in the vhost API. Add a WARNING.
Suggested-by: Joao Martins
Signed-off-by: Dongli Zhang
---
drivers/vhost/vhost.c | 18
Log write descriptors for the I/O queue, leveraging vhost_scsi_get_desc()
and vhost_get_vq_desc() to retrieve the array of write descriptors to
obtain the log buffer.
In addition, introduce a vhost-scsi specific function to log vring
descriptors. In this function, the 'partial' argument is set to
Adjust vhost_scsi_get_desc() to facilitate logging of vring descriptors.
Add new arguments to allow passing the log buffer and length to
vhost_get_vq_desc().
In addition, reset 'log_num' since vhost_get_vq_desc() may reset it only
after certain condition checks.
Suggested-by: Joao Martins
Signe
Log write descriptors for the control queue, leveraging
vhost_scsi_get_desc() and vhost_get_vq_desc() to retrieve the array of
write descriptors to obtain the log buffer.
For Task Management Requests, similar to the I/O queue, store the log
buffer during the submission path and log it in the compl
Although the support of VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 was
signaled by the commit 664ed90e621c ("vhost/scsi: Set
VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 feature bits"),
vhost_scsi_send_bad_target() still assumes the response in a single
descriptor.
Similar issue in vhost_scsi_send_bad_t
The live migration with vhost-scsi has been enabled by QEMU commit
b3e89c941a85 ("vhost-scsi: Allow user to enable migration"), which
thoroughly explains the workflow that QEMU collaborates with vhost-scsi on
the live migration.
Although it logs dirty data for the used ring, it doesn't log any wri
The vhost-scsi completion path may access vq->log_base when vq->log_used is
already set to false.
vhost-thread QEMU-thread
vhost_scsi_complete_cmd_work()
-> vhost_add_used()
-> vhost_add_used_n()
if (unlikely(vq->log_used))
Drivers with "memory-region" properties currently have to do their own
parsing of "memory-region" properties. The result is all the drivers
have similar patterns of a call to parse "memory-region" and then get
the region's address and size. As this is a standard property, it should
have common func
While there's a common function to parse "memory-region" properties for
DMA pool regions, there's not anything for driver private regions. As a
result, drivers have resorted to parsing "memory-region" properties
themselves repeating the same pattern over and over. To fix this, this
series adds 2 fu
On Mon, 17 Mar 2025 15:33:01 +0200, Daniel Baluta wrote:
> DSP found in i.MX8MP SOC can be used by multiple frameworks in order to
> enable various applications:
> - rproc/rpmsg framework, used to load for example Zephyr samples
> - Sound Open Firmware, used to enable various audi
On Mon, Mar 17, 2025 at 4:08 AM Vlastimil Babka wrote:
>
> On 3/14/25 18:10, Suren Baghdasaryan wrote:
> > On Tue, Mar 4, 2025 at 11:08 AM Liam R. Howlett
> > wrote:
> >>
> >> * Vlastimil Babka [250304 05:55]:
> >> > On 2/25/25 21:26, Suren Baghdasaryan wrote:
> >> > > On Mon, Feb 24, 2025 at 1
Hello Stafford and Geert,
Thank you for the reviews.
On 3/17/25 1:55 PM, Geert Uytterhoeven wrote:
On Sun, 16 Mar 2025 at 07:59, Stafford Horne wrote:
[...]
+struct cache_desc {
+ u32 size;
+ u32 sets;
+ u32 block_size;
+ u32 ways;
Considering the changes below to add cache
On Mon, Mar 17, 2025 at 06:52:57PM +0100, Thomas Weißschuh wrote:
> On 2025-03-17 08:37:46+0100, Willy Tarreau wrote:
> > On Sun, Mar 16, 2025 at 02:55:02PM +0100, Thomas Weißschuh wrote:
> > > Add support for 32bit and 64bit SPARC to nolibc.
> >
> > Oh nice!
> >
> > > Signed-off-by: Thomas Weißs
On 2025-03-17 08:37:46+0100, Willy Tarreau wrote:
> On Sun, Mar 16, 2025 at 02:55:02PM +0100, Thomas Weißschuh wrote:
> > Add support for 32bit and 64bit SPARC to nolibc.
>
> Oh nice!
>
> > Signed-off-by: Thomas Weißschuh
> > ---
> > This is only tested on QEMU.
> > Any tests on real hardware wo
From: Vladis Dronov
[ Upstream commit 65be5c95d08eedda570a6c888a12384c77fe7614 ]
The kernel requires X86_FEATURE_SGX_LC to be able to create SGX enclaves,
not just X86_FEATURE_SGX.
There is quite a number of hardware which has X86_FEATURE_SGX but not
X86_FEATURE_SGX_LC. A kernel running on such
On 6/19/24 11:02 PM, Luca Weiss wrote:
Add the mailbox compatible for MSM8226 SoC.
Hi all,
This patch is still pending.
Regards
Luca
Signed-off-by: Luca Weiss
---
Documentation/devicetree/bindings/mailbox/qcom,apcs-kpss-global.yaml | 1 +
1 file changed, 1 insertion(+)
diff --git
a/D
On 10/19/24 12:06 PM, Luca Weiss wrote:
On Donnerstag, 6. Juni 2024 21:01:36 MESZ Luca Weiss wrote:
When the mailbox driver has not probed yet, the error message "failed to
parse smd edge" is just going to confuse users, so improve the error
prints a bit.
Cover the last remaining exits from qco
On Mon, Mar 17, 2025 at 12:47 AM wrote:
>
> From: FengWei
>
> strncpy() is deprecated for NUL-terminated destination buffers. Use
> strscpy() instead and remove the manual NUL-termination.
>
> Signed-off-by: FengWei
> ---
> tools/testing/selftests/bpf/test_verifier.c | 3 +--
> 1 file changed,
From: Vladis Dronov
[ Upstream commit 65be5c95d08eedda570a6c888a12384c77fe7614 ]
The kernel requires X86_FEATURE_SGX_LC to be able to create SGX enclaves,
not just X86_FEATURE_SGX.
There is quite a number of hardware which has X86_FEATURE_SGX but not
X86_FEATURE_SGX_LC. A kernel running on such
From: Vladis Dronov
[ Upstream commit 65be5c95d08eedda570a6c888a12384c77fe7614 ]
The kernel requires X86_FEATURE_SGX_LC to be able to create SGX enclaves,
not just X86_FEATURE_SGX.
There is quite a number of hardware which has X86_FEATURE_SGX but not
X86_FEATURE_SGX_LC. A kernel running on such
From: Vladis Dronov
[ Upstream commit 65be5c95d08eedda570a6c888a12384c77fe7614 ]
The kernel requires X86_FEATURE_SGX_LC to be able to create SGX enclaves,
not just X86_FEATURE_SGX.
There is quite a number of hardware which has X86_FEATURE_SGX but not
X86_FEATURE_SGX_LC. A kernel running on such
From: Florian Westphal
[ Upstream commit c21b02fd9cbf15aed6e32c89e0fd70070281e3d1 ]
These scripts fail if the kernel is tainted which leads to wrong test
failure reports in CI environments when an unrelated test triggers some
splat.
Check taint state at start of script and SKIP if its already d
From: Vladis Dronov
[ Upstream commit 65be5c95d08eedda570a6c888a12384c77fe7614 ]
The kernel requires X86_FEATURE_SGX_LC to be able to create SGX enclaves,
not just X86_FEATURE_SGX.
There is quite a number of hardware which has X86_FEATURE_SGX but not
X86_FEATURE_SGX_LC. A kernel running on such
From: Florian Westphal
[ Upstream commit c21b02fd9cbf15aed6e32c89e0fd70070281e3d1 ]
These scripts fail if the kernel is tainted which leads to wrong test
failure reports in CI environments when an unrelated test triggers some
splat.
Check taint state at start of script and SKIP if its already d
On Fri, 14 Mar 2025 at 06:37, David Gow wrote:
>
> On Fri, 14 Mar 2025 at 03:27, Rae Moar wrote:
> >
> > A bug was identified where the KTAP below caused an infinite loop:
> >
> > TAP version 13
> > ok 4 test_case
> > 1..4
> >
> > The infinite loop was caused by the parser not parsing a test p
On Mon, Mar 17, 2025 at 5:30 PM Alexander Stein
wrote:
>
> Am Montag, 17. März 2025, 14:33:03 CET schrieb Daniel Baluta:
> > Enable MU2 node and add mu2 root clock.
> > MU2 is used to communicate with DSP core.
> >
> > Reviewed-by: Iuliana Prodan
> > Reviewed-by: Peng Fan
> > Signed-off-by: Dani
Am Montag, 17. März 2025, 14:33:03 CET schrieb Daniel Baluta:
> Enable MU2 node and add mu2 root clock.
> MU2 is used to communicate with DSP core.
>
> Reviewed-by: Iuliana Prodan
> Reviewed-by: Peng Fan
> Signed-off-by: Daniel Baluta
> ---
> arch/arm64/boot/dts/freescale/imx8mp.dtsi | 2 +-
>
Enable MU2 node and add mu2 root clock.
MU2 is used to communicate with DSP core.
Reviewed-by: Iuliana Prodan
Reviewed-by: Peng Fan
Signed-off-by: Daniel Baluta
---
arch/arm64/boot/dts/freescale/imx8mp.dtsi | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/arm64/boot/dts
From: "Liam R. Howlett"
Make testing work for the slab and rcu changes that have come in with
the sheaves work.
This only works with one kmem_cache, and only the first one used.
Subsequent setting of kmem_cache will not update the active kmem_cache
and will be silently dropped because there are
Setup the maple_node_cache with percpu sheaves of size 32 to hopefully
improve its performance. Change the single node rcu freeing in
ma_free_rcu() to use kfree_rcu() instead of the custom callback, which
allows the rcu_free sheaf batching to be used. Note there are other
users of mt_free_rcu() whe
Add functions for efficient guaranteed allocations e.g. in a critical
section that cannot sleep, when the exact number of allocations is not
known beforehand, but an upper limit can be calculated.
kmem_cache_prefill_sheaf() returns a sheaf containing at least given
number of objects.
kmem_cache_a
From: "Liam R. Howlett"
Allocate a sheaf and fill it to the count amount. Does not fill to the
sheaf limit to detect incorrect allocation requests.
Signed-off-by: Liam R. Howlett
---
tools/include/linux/slab.h | 24 +
tools/testing/shared/linux.c | 84 +++
Extend the sheaf infrastructure for more efficient kfree_rcu() handling.
For caches with sheaves, on each cpu maintain a rcu_free sheaf in
addition to main and spare sheaves.
kfree_rcu() operations will try to put objects on this sheaf. Once full,
the sheaf is detached and submitted to call_rcu()
Specifying a non-zero value for a new struct kmem_cache_args field
sheaf_capacity will setup a caching layer of percpu arrays called
sheaves of given capacity for the created cache.
Allocations from the cache will allocate via the percpu sheaves (main or
spare) as long as they have no NUMA node pr
The possibility of many barn operations is determined by the current
number of full or empty sheaves. Taking the barn->lock just to find out
that e.g. there are no empty sheaves results in unnecessary overhead and
lock contention. Thus perform these checks outside of the lock with a
data_race() ann
From: Sebastian Andrzej Siewior
In !PREEMPT_RT local_lock_irqsave() disables interrupts to protect
critical section, but it doesn't prevent NMI, so the fully reentrant
code cannot use local_lock_irqsave() for exclusive access.
Introduce localtry_lock_t and localtry_lock_irqsave() that
disables i
Hi,
This is the v3 RFC to add an opt-in percpu array-based caching layer to
SLUB. This is to publish accumulated fixes since v2 ahead of LSF/MM. I
have also squashed the usage of localtry_lock to be used immediately
(not converted to in a separate patch) as the lock should be going to
6.15 at this
DSP core needs ocram, core and debug clocks.
Reviewed-by: Iuliana Prodan
Reviewed-by: Peng Fan
Signed-off-by: Daniel Baluta
---
arch/arm64/boot/dts/freescale/imx8mp.dtsi | 5 +
1 file changed, 5 insertions(+)
diff --git a/arch/arm64/boot/dts/freescale/imx8mp.dtsi
b/arch/arm64/boot/dts/fr
DSP can be used with various frameworks (e.g audio firmware, rproc).
Currently 'dsp' configuration is intended for audio firmware but it
doesn't work well with board level DTs (e.g imx8mp-evk) because
board level DT enables audio related IPs (e.g SAI) while audio firmware
needs this IPs disabled (b
Add resets property to dsp node in order to be able to control the dsp
run/stall bit from audio block control.
Reviewed-by: Peng Fan
Reviewed-by: Frank Li
Signed-off-by: Daniel Baluta
---
arch/arm64/boot/dts/freescale/imx8mp.dtsi | 3 +++
1 file changed, 3 insertions(+)
diff --git a/arch/arm6
With imx8mp-evk board we are now configuring 'dsp' node for rproc usage,
so add rproc specific memory regions.
Also, enable dsp node because it is ready to be used.
Reviewed-by: Ahmad Fatoum
Reviewed-by: Iuliana Prodan
Reviewed-by: Peng Fan
Signed-off-by: Daniel Baluta
---
arch/arm64/boot/dt
On 17/03/2025 13:47, Sabrina Dubroca wrote:
Hello,
A few comments since it seems you'll have to send one more version
(otherwise they could be fixed later).
2025-03-12, 21:54:27 +0100, Antonio Quartulli wrote:
diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c
index
8d267d4c
The k3_{r5/dsp/m4}_reserved_mem_init() functions initialize the reserved
memory regions associated with the remote processor. Refactor these
functions into ti_k3_common.c driver as k3_reserved_mem_init() and align
R5, DSP and M4 drivers to use this common function throughout.
Signed-off-by: Belesw
Hello,
A few comments since it seems you'll have to send one more version
(otherwise they could be fixed later).
2025-03-12, 21:54:27 +0100, Antonio Quartulli wrote:
> diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c
> index
> 8d267d4c82283d9b5f989478102086ce385195d5..407b5b9
The ti_k3_m4_remoteproc.c driver previously hardcoded device memory
region addresses and names. Change this to use the k3_rproc_mem_data
structure to store memory information. This aligns with DSP and R5
drivers, and can be refactored out later.
Signed-off-by: Beleswar Padhi
---
drivers/remotepr
The k3_{r5/dsp/m4}_release_tsp() functions release the TI-SCI processor
control of a remote processor, which is auto triggered upon device
removal. Refactor these functions into ti_k3_common.c driver as
k3_release_tsp() and align R5, DSP and M4 drivers to use this common
function throughout.
Signe
The k3_r5_core_of_get_internal_memories() and
k3_{/dsp/m4}_rproc_of_get_memories() functions initialize and assigns
memory regions used by the remote processor. Refactor these
implementations into ti_k3_common.c driver as k3_rproc_of_get_memories()
and align R5, DSP and M4 drivers to use this commo
The ti_k3_r5_remoteproc.c driver previously hardcoded device memory
region addresses and names. Change this to use the k3_r5_rproc_mem_data
structure to store memory information. This aligns with K3 DSP and M4
drivers, and can be refactored out later.
Signed-off-by: Beleswar Padhi
---
drivers/re
Hi Konrad,
On Fri Mar 14, 2025 at 11:08 PM CET, Konrad Dybcio wrote:
> On 3/14/25 10:17 AM, Luca Weiss wrote:
>> UFS host controller, when scaling gears, should choose appropriate
>> performance state of RPMh power domain controller along with clock
>> frequency. So let's add the OPP table support
The k3_{r5/dsp}_mem_release() and k3_m4_rproc_dev_mem_release()
functions release the reserved memory of the device, which get auto
triggered upon device removal. Refactor these functions into
ti_k3_common.c driver as k3_mem_release() and align R5, DSP and M4
drivers to use this common function thr
The .da_to_va rproc ops implementations in TI K3 R5, DSP and M4
remoteproc drivers return the Kernel virtual address for a
corresponding rproc device address. Refactor the implementations into
ti_k3_common.c driver as k3_rproc_da_to_va and align R5, DSP and M4
drivers to use this common function fo
The .attach rproc ops implementations in TI K3 R5, DSP and M4 drivers
are NOPs. Refactor the implementations into ti_k3_common.c driver
as k3_rproc_attach() and align R5, DSP and M4 drivers to register this
common function as .attach ops.
Signed-off-by: Beleswar Padhi
---
drivers/remoteproc/ti_k
The .get_loaded_rsc_table rproc ops implementations in TI K3 R5, DSP and
M4 remoteproc drivers return a pointer to the resource table that was
pre-loaded at the base address of the DDR region reserved for firmware
usage. Refactor the implementations into ti_k3_common.c driver as
k3_get_loaded_rsc_t
The .detach rproc ops implementations in TI K3 R5, DSP and M4
remoteproc drivers are NOPs. Refactor the implementations into
ti_k3_common.c driver as k3_rproc_detach() and align R5, DSP and M4
drivers to register this common function as .detach ops.
Signed-off-by: Beleswar Padhi
---
drivers/remo
The k3_dsp_rproc_release() function erroneously deasserts the local
reset even for devices which do not support it. Even though it results
in a no-operation, Update the logic to explicitly deassert the local
reset for devices that support it and only the global reset for those
that do not.
Further
The .stop rproc ops implementations in ti_k3_dsp_remoteproc.c and
ti_k3_m4_remoteproc.c drivers put the remote processor into reset.
Refactor the implementations into ti_k3_common.c driver as
k3_rproc_stop() and align DSP and M4 drivers to register this common
function as .stop ops.
Further, do no
The .unprepare rproc ops implementations in ti_k3_dsp_remoteproc.c and
ti_k3_m4_remoteproc.c drivers assert the global reset on the remote
processor. Refactor the implementations into ti_k3_common.c driver as
k3_rproc_unprepare() and align DSP and M4 drivers to register this
common function as .unp
The k3_{r5/dsp}_rproc_request_mbox() function acquires the mailbox
channel and sends a message through the acquired channel. The TI K3 M4
remoteproc driver acquires the mailbox channel in probe and sends the
message later in .attach()/.start() callbacks. Refactor the
k3_{r5/dsp}_rproc_request_mbox(
The k3_dsp_rproc_start() function sets the boot address and releases the
reset on the remote processor. Whereas, the k3_m4_rproc_start() function
only needs to release the reset. Refactor the k3_m4_rproc_start() into
ti_k3_common.c as k3_rproc_start() and align the DSP and M4 drivers to
invoke this
The k3_dsp_rproc_reset() function erroneously asserts the local reset
even for devices which do not support it. Even though it results in a
no-operation, Update the logic to explicitly assert the local reset for
devices that support it and only the global reset for those that do not.
Further, refa
The existing implementation of the waiting mechanism in
"k3_r5_cluster_rproc_init()" waits for the "released_from_reset" flag to
be set as part of the firmware boot process in "k3_r5_rproc_start()".
The "k3_r5_cluster_rproc_init()" function is invoked in the probe
routine which causes unexpected fa
The ti_k3_dsp_remoteproc.c driver asserts the local reset in probe and
releases the global reset in .prepare callback. Whereas, the
ti_k3_m4_remoteproc.c driver does both operations in .prepare callback,
which is more suitable as it ensures lreset is asserted for subsequent
core start operations fr
From: Siddharth Vadapalli
Commit ea1d6fb5b571 ("remoteproc: k3-dsp: Acquire mailbox handle during
probe routine") introduced a check in the "k3_dsp_rproc_mbox_callback()"
and "k3_dsp_rproc_kick()" callbacks, causing them to exit if the remote
core's state is "RPROC_DETACHED". However, the "__rpro
From: Siddharth Vadapalli
Commit f3f11cfe8907 ("remoteproc: k3-r5: Acquire mailbox handle during
probe routine") introduced a check in the "k3_r5_rproc_mbox_callback()"
and "k3_r5_rproc_kick()" callbacks, causing them to exit if the remote
core's state is "RPROC_DETACHED". However, the "__rproc_a
The mailbox .rx_callback implementations in TI K3 R5, DSP and M4
remoteproc drivers handle inbound mailbox messages in the same way.
Introduce a common driver 'ti_k3_common.c' and refactor the
implementations into a common function 'k3_rproc_mbox_callback'() in it.
Signed-off-by: Beleswar Padhi
-
The .kick rproc ops implementations in TI K3 R5, DSP and M4 remoteproc
drivers sends a mailbox message to the remote processor in the same
way. Refactor the implementations into a common function
'k3_rproc_kick()' in the ti_k3_common.c driver.
Signed-off-by: Beleswar Padhi
---
drivers/remoteproc
1 - 100 of 136 matches
Mail list logo
