[net-next v41] mctp pcc: Implement MCTP over PCC Transport

[net-next v41] mctp pcc: Implement MCTP over PCC Transport

[Adam Young]

Implementation of network driver for
Management Component Transport Protocol(MCTP)
over Platform Communication Channel(PCC)

DMTF DSP:0292
Link: https://www.dmtf.org/sites/default/files/standards/documents/DSP0292_1.0.0WIP50.pdf

The transport mechanism is called Platform Communication Channels (PCC)
is part of the ACPI spec:

Link: https://uefi.org/htmlspecs/ACPI_Spec_6_4_html/14_Platform_Communications_Channel/Platform_Comm_Channel.html

The PCC mechanism is managed via a mailbox implemented at
drivers/mailbox/pcc.c

MCTP devices are specified via ACPI by entries in DSDT/SSDT and
reference channels specified in the PCCT. Messages are sent on a type
3 and received on a type 4 channel.  Communication with other devices
use the PCC based doorbell mechanism; a shared memory segment with a
corresponding interrupt and a memory register used to trigger remote
interrupts.

The shared buffer must be at least 68 bytes long as that is the minimum
MTU as defined by the MCTP specification.

Unlike the existing PCC Type 2 based drivers, the mssg parameter to
mbox_send_msg is actively used. The data section of the struct sk_buff
that contains the outgoing packet is sent to the mailbox, already
properly formatted as a PCC exctended message.

If the mailbox ring buffer is full, the driver stops the incoming
packet queues until a message has been sent, freeing space in the
ring buffer.

When the Type 3 channel outbox receives a txdone response interrupt,
it consumes the outgoing sk_buff, allowing it to be freed.

Bringing up an interface creates the channel between the network driver
and the mailbox driver. This enables communication with the remote
endpoint, to include the receipt of new messages. Bringing down an
interface removes the channel, and no new messages can be delivered.
Stopping the interface will leave any packets that are cached in the
mailbox ringbuffer. They cannot safely be freed until the PCC mailbox
attempts to deliver them and has removed them from the ring buffer.

PCC is based on a shared buffer and a set of I/O mapped memory locations
that the Spec calls registers.  This mechanism exists regardless of the
existence of the driver. If the user has the ability to map these
physical location to virtual locations, they have the ability to drive the
hardware.  Thus, there is a security aspect to this mechanism that extends
beyond the responsibilities of the operating system.

If the hardware does not expose the PCC in the ACPI table, this device
will never be enabled. Thus it is only an issue on hardware that does
support PCC. In that case, it is up to the remote controller to sanitize
communication; MCTP will be exposed as a socket interface, and userland
can send any crafted packet it wants. It would also be incumbent on
the hardware manufacturer to allow the end user to disable MCTP over PCC
communication if they did not want to expose it.

Link: https://www.dmtf.org/sites/default/files/standards/documents/DSP0292_1.0.0WIP50.pdf
Link: https://uefi.org/htmlspecs/ACPI_Spec_6_4_html/14_Platform_Communications_Channel/Platform_Comm_Channel.html
Signed-off-by: Adam Young <admiyo@os.amperecomputing.com>

---

Previous Version:
https://lore.kernel.org/lkml/20260508032953.337036-1-admiyo@os.amperecomputing.com/

Changes from Previous version

Remove check for skb_is_nonlinear(skb) as it is done in skb_linearize(skb)
Removed comment about BE support
Spacing changes after gotos
Moved comment on IRQ unsafe stats from tx_done to tx_prepare
---
 MAINTAINERS                 |   5 +
 drivers/net/mctp/Kconfig    |  16 ++
 drivers/net/mctp/Makefile   |   1 +
 drivers/net/mctp/mctp-pcc.c | 410 ++++++++++++++++++++++++++++++++++++
 4 files changed, 432 insertions(+)
 create mode 100644 drivers/net/mctp/mctp-pcc.c

diff --git a/MAINTAINERS b/MAINTAINERS
index 5bbbbde6b907..252c77b24791 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15403,6 +15403,11 @@ F:	include/net/mctpdevice.h
 F:	include/net/netns/mctp.h
 F:	net/mctp/
 
+MANAGEMENT COMPONENT TRANSPORT PROTOCOL (MCTP) over PCC (MCTP-PCC) Driver
+M:	Adam Young <admiyo@os.amperecomputing.com>
+S:	Maintained
+F:	drivers/net/mctp/mctp-pcc.c
+
 MAPLE TREE
 M:	Liam R. Howlett <liam@infradead.org>
 R:	Alice Ryhl <aliceryhl@google.com>
diff --git a/drivers/net/mctp/Kconfig b/drivers/net/mctp/Kconfig
index cf325ab0b1ef..e68d23794a80 100644
--- a/drivers/net/mctp/Kconfig
+++ b/drivers/net/mctp/Kconfig
@@ -47,6 +47,22 @@ config MCTP_TRANSPORT_I3C
 	  A MCTP protocol network device is created for each I3C bus
 	  having a "mctp-controller" devicetree property.
 
+config MCTP_TRANSPORT_PCC
+	tristate "MCTP PCC transport"
+	depends on ACPI
+	depends on PCC
+	depends on 64BIT
+	depends on CPU_LITTLE_ENDIAN
+	help
+	  Provides a driver to access MCTP devices over PCC transport,
+	  A MCTP protocol network device is created via ACPI for each
+	  entry in the DSDT/SSDT that matches the identifier. The Platform
+	  communication channels are selected from the corresponding
+	  entries in the PCCT.
+
+	  Say y here if you need to connect to MCTP endpoints over PCC. To
+	  compile as a module, use m; the module will be called mctp-pcc.
+
 config MCTP_TRANSPORT_USB
 	tristate "MCTP USB transport"
 	depends on USB
diff --git a/drivers/net/mctp/Makefile b/drivers/net/mctp/Makefile
index c36006849a1e..0a591299ffa9 100644
--- a/drivers/net/mctp/Makefile
+++ b/drivers/net/mctp/Makefile
@@ -1,4 +1,5 @@
 obj-$(CONFIG_MCTP_SERIAL) += mctp-serial.o
 obj-$(CONFIG_MCTP_TRANSPORT_I2C) += mctp-i2c.o
 obj-$(CONFIG_MCTP_TRANSPORT_I3C) += mctp-i3c.o
+obj-$(CONFIG_MCTP_TRANSPORT_PCC) += mctp-pcc.o
 obj-$(CONFIG_MCTP_TRANSPORT_USB) += mctp-usb.o
diff --git a/drivers/net/mctp/mctp-pcc.c b/drivers/net/mctp/mctp-pcc.c
new file mode 100644
index 000000000000..a9962c67911c
--- /dev/null
+++ b/drivers/net/mctp/mctp-pcc.c
@@ -0,0 +1,410 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * mctp-pcc.c - Driver for MCTP over PCC.
+ * Copyright (c) 2024-2026, Ampere Computing LLC
+ *
+ */
+
+/* Implementation of MCTP over PCC DMTF Specification DSP0256
+ * https://www.dmtf.org/sites/default/files/standards/documents/DSP0292_1.0.0WIP50.pdf
+ */
+
+#include <linux/acpi.h>
+#include <linux/hrtimer.h>
+#include <linux/if_arp.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/mailbox_client.h>
+#include <linux/module.h>
+#include <linux/netdevice.h>
+#include <linux/platform_device.h>
+#include <linux/skbuff.h>
+#include <linux/string.h>
+
+#include <acpi/acpi_bus.h>
+#include <acpi/acpi_drivers.h>
+#include <acpi/acrestyp.h>
+#include <acpi/actbl.h>
+#include <acpi/pcc.h>
+#include <net/mctp.h>
+#include <net/mctpdevice.h>
+
+#define MCTP_SIGNATURE          "MCTP"
+#define MCTP_SIGNATURE_LENGTH   (sizeof(MCTP_SIGNATURE) - 1)
+#define MCTP_MIN_MTU            68
+#define PCC_HEADER_SIZE         sizeof(struct acpi_pcct_ext_pcc_shared_memory)
+#define MCTP_PCC_MIN_SIZE       (PCC_HEADER_SIZE + MCTP_MIN_MTU)
+#define PCC_EXTRA_LEN           (PCC_HEADER_SIZE - sizeof(pcc_header.command))
+struct mctp_pcc_mailbox {
+	u32 index;
+	struct pcc_mbox_chan *chan;
+	struct mbox_client client;
+};
+
+/* The netdev structure. One of these per PCC adapter. */
+struct mctp_pcc_ndev {
+	struct net_device *ndev;
+	struct acpi_device *acpi_device;
+	struct mctp_pcc_mailbox inbox;
+	struct mctp_pcc_mailbox outbox;
+};
+
+static void mctp_pcc_client_rx_callback(struct mbox_client *cl, void *mssg)
+{
+	struct acpi_pcct_ext_pcc_shared_memory pcc_header;
+	struct mctp_pcc_ndev *mctp_pcc_ndev;
+	struct mctp_pcc_mailbox *inbox;
+	struct mctp_skb_cb *cb;
+	struct sk_buff *skb;
+	u32 header_length;
+	int size;
+
+	mctp_pcc_ndev = container_of(cl, struct mctp_pcc_ndev, inbox.client);
+	inbox = &mctp_pcc_ndev->inbox;
+	memcpy_fromio(&pcc_header, inbox->chan->shmem, sizeof(pcc_header));
+
+	// The message must at least have the PCC command indicating it is an MCTP
+	// message followed by the MCTP header, or we have a malformed message.
+	// This may be run on big endian system, but the data in the buffer is
+	// explicitly little endian.
+	header_length = le32_to_cpu(pcc_header.length);
+
+	if (header_length < sizeof(pcc_header.command) + sizeof(struct mctp_hdr))
+		goto error;
+
+	// If the reported size is larger than the shared memory minus headers,
+	// something is wrong and treat the buffer as corrupted data.
+	if (header_length > inbox->chan->shmem_size - PCC_EXTRA_LEN)
+		goto error;
+
+	if (memcmp(&pcc_header.command, MCTP_SIGNATURE, MCTP_SIGNATURE_LENGTH) != 0)
+		goto error;
+
+	size = header_length + PCC_EXTRA_LEN;
+	skb = netdev_alloc_skb(mctp_pcc_ndev->ndev, size);
+	if (!skb)
+		goto error;
+	skb_put(skb, size);
+	skb->protocol = htons(ETH_P_MCTP);
+	memcpy_fromio(skb->data, inbox->chan->shmem, size);
+	dev_dstats_rx_add(mctp_pcc_ndev->ndev, size);
+	skb_pull(skb, sizeof(pcc_header));
+	skb_reset_mac_header(skb);
+	skb_reset_network_header(skb);
+	cb = __mctp_cb(skb);
+	cb->halen = 0;
+	netif_rx(skb);
+	return;
+error:
+	dev_dstats_rx_dropped(mctp_pcc_ndev->ndev);
+}
+
+static netdev_tx_t mctp_pcc_tx(struct sk_buff *skb, struct net_device *ndev)
+{
+	struct acpi_pcct_ext_pcc_shared_memory *pcc_header;
+	struct mctp_pcc_ndev *mpnd = netdev_priv(ndev);
+	int len = skb->len;
+
+	/* Consolidated a fragmented packet into contiguous memory */
+	if (skb_linearize(skb))
+		goto error;
+
+	if (skb_cow_head(skb, sizeof(*pcc_header)))
+		goto error;
+	pcc_header = skb_push(skb, sizeof(*pcc_header));
+	pcc_header->signature = PCC_SIGNATURE | mpnd->outbox.index;
+	pcc_header->flags = PCC_CMD_COMPLETION_NOTIFY;
+	memcpy(&pcc_header->command, MCTP_SIGNATURE, MCTP_SIGNATURE_LENGTH);
+	pcc_header->length = len + MCTP_SIGNATURE_LENGTH;
+
+	if (mbox_send_message(mpnd->outbox.chan->mchan, skb) < 0) {
+		// Remove the header in case it gets sent again
+		skb_pull(skb, sizeof(*pcc_header));
+		netif_stop_queue(ndev);
+		return NETDEV_TX_BUSY;
+	}
+
+	return NETDEV_TX_OK;
+error:
+	dev_dstats_tx_dropped(ndev);
+	kfree_skb(skb);
+	return NETDEV_TX_OK;
+}
+
+static void mctp_pcc_tx_prepare(struct mbox_client *cl, void *mssg)
+{
+	struct mctp_pcc_ndev *mctp_pcc_ndev;
+	struct mctp_pcc_mailbox *outbox;
+	struct sk_buff *skb = mssg;
+
+	mctp_pcc_ndev = container_of(cl, struct mctp_pcc_ndev, outbox.client);
+	outbox = &mctp_pcc_ndev->outbox;
+
+	/* The PCC Mailbox typically does not make use of the mssg pointer
+	 * The mctp-over pcc driver is the only client that uses it.
+	 * This value should always be non-null; it is possible
+	 * that a change in the Mailbox level will break that assumption.
+	 */
+	if (!skb) {
+		netdev_warn_once(mctp_pcc_ndev->ndev,
+				 "%s called with null message.\n", __func__);
+		return;
+	}
+
+	if (skb->len > outbox->chan->shmem_size) {
+		dev_dstats_tx_dropped(mctp_pcc_ndev->ndev);
+		return;
+	}
+	memcpy_toio(outbox->chan->shmem, skb->data, skb->len);
+
+	/*
+	 * This packet could still be dropped in the PCC layer,
+	 * But the only place that could deal with that is mctp_pcc_tx_done.
+	 * That is called from the Hard-IRQ handler, and it is not safe to
+	 * call stats functions from HARD-IRQ context.
+	 */
+
+	dev_dstats_tx_add(mctp_pcc_ndev->ndev, skb->len);
+}
+
+static void mctp_pcc_tx_done(struct mbox_client *c, void *mssg, int r)
+{
+	struct mctp_pcc_ndev *mctp_pcc_ndev;
+	struct sk_buff *skb = mssg;
+
+	/*
+	 * If there is a packet in flight during driver cleanup
+	 * It may have been freed already.
+	 */
+	if (!mssg)
+		return;
+	mctp_pcc_ndev = container_of(c, struct mctp_pcc_ndev, outbox.client);
+	dev_consume_skb_any(skb);
+	netif_wake_queue(mctp_pcc_ndev->ndev);
+}
+
+static int mctp_pcc_open(struct net_device *ndev)
+{
+	struct mctp_pcc_ndev *mctp_pcc_ndev = netdev_priv(ndev);
+	struct mctp_pcc_mailbox *outbox, *inbox;
+
+	outbox = &mctp_pcc_ndev->outbox;
+	inbox = &mctp_pcc_ndev->inbox;
+
+	outbox->chan = pcc_mbox_request_channel(&outbox->client, outbox->index);
+	if (IS_ERR(outbox->chan))
+		return PTR_ERR(outbox->chan);
+	if (outbox->chan->shmem_size < MCTP_PCC_MIN_SIZE) {
+		pcc_mbox_free_channel(outbox->chan);
+		return -EINVAL;
+	}
+
+	inbox->client.rx_callback = mctp_pcc_client_rx_callback;
+	inbox->chan = pcc_mbox_request_channel(&inbox->client, inbox->index);
+	if (IS_ERR(inbox->chan)) {
+		pcc_mbox_free_channel(outbox->chan);
+		return PTR_ERR(inbox->chan);
+	}
+	if (inbox->chan->shmem_size < MCTP_PCC_MIN_SIZE) {
+		pcc_mbox_free_channel(outbox->chan);
+		pcc_mbox_free_channel(inbox->chan);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int mctp_pcc_stop(struct net_device *ndev)
+{
+	struct mctp_pcc_ndev *mctp_pcc_ndev;
+	unsigned int count, idx;
+	struct mbox_chan *chan;
+	struct sk_buff *skb;
+
+	mctp_pcc_ndev = netdev_priv(ndev);
+	chan = mctp_pcc_ndev->outbox.chan->mchan;
+	pcc_mbox_free_channel(mctp_pcc_ndev->inbox.chan);
+	scoped_guard(spinlock_irqsave, &chan->lock) {
+		skb = chan->active_req;
+		chan->active_req = NULL;
+		if (skb) {
+			dev_dstats_tx_dropped(ndev);
+			dev_consume_skb_any(skb);
+		}
+		while (chan->msg_count > 0) {
+			count = chan->msg_count;
+			idx = chan->msg_free;
+			if (idx >= count)
+				idx -= count;
+			else
+				idx += MBOX_TX_QUEUE_LEN - count;
+			skb = chan->msg_data[idx];
+			dev_dstats_tx_dropped(ndev);
+			dev_consume_skb_any(skb);
+			chan->msg_count--;
+		}
+	}
+	pcc_mbox_free_channel(mctp_pcc_ndev->outbox.chan);
+	return 0;
+}
+
+static const struct net_device_ops mctp_pcc_netdev_ops = {
+	.ndo_open = mctp_pcc_open,
+	.ndo_stop = mctp_pcc_stop,
+	.ndo_start_xmit = mctp_pcc_tx,
+};
+
+static void mctp_pcc_setup(struct net_device *ndev)
+{
+	ndev->type = ARPHRD_MCTP;
+	ndev->hard_header_len = sizeof(struct acpi_pcct_ext_pcc_shared_memory);
+	ndev->tx_queue_len = 0;
+	ndev->flags = IFF_NOARP;
+	ndev->netdev_ops = &mctp_pcc_netdev_ops;
+	ndev->needs_free_netdev = true;
+	ndev->pcpu_stat_type = NETDEV_PCPU_STAT_DSTATS;
+}
+
+struct mctp_pcc_lookup_context {
+	int index;
+	u32 inbox_index;
+	u32 outbox_index;
+};
+
+static acpi_status lookup_pcct_indices(struct acpi_resource *ares,
+				       void *context)
+{
+	struct mctp_pcc_lookup_context *luc = context;
+	struct acpi_resource_address32 *addr;
+
+	if (ares->type != ACPI_RESOURCE_TYPE_ADDRESS32)
+		return AE_OK;
+
+	addr = ACPI_CAST_PTR(struct acpi_resource_address32, &ares->data);
+	switch (luc->index) {
+	case 0:
+		luc->outbox_index = addr[0].address.minimum;
+		break;
+	case 1:
+		luc->inbox_index = addr[0].address.minimum;
+		break;
+	default:
+		return AE_ERROR;
+	}
+	luc->index++;
+	return AE_OK;
+}
+
+static void mctp_cleanup_netdev(void *data)
+{
+	struct net_device *ndev = data;
+
+	mctp_unregister_netdev(ndev);
+}
+
+static int initialize_mtu(struct net_device *ndev)
+{
+	struct mctp_pcc_ndev *mctp_pcc_ndev;
+	struct mctp_pcc_mailbox *outbox;
+	struct pcc_mbox_chan *pchan;
+	int mctp_pcc_max_mtu;
+
+	mctp_pcc_ndev = netdev_priv(ndev);
+	outbox = &mctp_pcc_ndev->outbox;
+	pchan = pcc_mbox_request_channel(&outbox->client, outbox->index);
+	if (IS_ERR(pchan))
+		return PTR_ERR(pchan);
+	if (pchan->shmem_size < MCTP_MIN_MTU + sizeof(struct acpi_pcct_ext_pcc_shared_memory)) {
+		pcc_mbox_free_channel(pchan);
+		return -EINVAL;
+	}
+	mctp_pcc_max_mtu = pchan->shmem_size - sizeof(struct acpi_pcct_ext_pcc_shared_memory);
+	pcc_mbox_free_channel(pchan);
+
+	ndev->mtu = MCTP_MIN_MTU;
+	ndev->max_mtu = mctp_pcc_max_mtu;
+	ndev->min_mtu = MCTP_MIN_MTU;
+
+	return 0;
+}
+
+static int mctp_pcc_driver_add(struct acpi_device *acpi_dev)
+{
+	struct mctp_pcc_lookup_context context = {0};
+	struct mctp_pcc_ndev *mctp_pcc_ndev;
+	struct device *dev = &acpi_dev->dev;
+	struct net_device *ndev;
+	acpi_handle dev_handle;
+	acpi_status status;
+	char name[32];
+	int rc;
+
+	dev_dbg(dev, "Adding mctp_pcc device for HID %s\n",
+		acpi_device_hid(acpi_dev));
+	dev_handle = acpi_device_handle(acpi_dev);
+	status = acpi_walk_resources(dev_handle, "_CRS", lookup_pcct_indices,
+				     &context);
+	if (!ACPI_SUCCESS(status)) {
+		dev_err(dev, "FAILED to lookup PCC indexes from CRS\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Ensure we have exactly 2 channels: an outbox and an inbox.
+	 */
+	if (context.index != 2)
+		return -EINVAL;
+
+	snprintf(name, sizeof(name), "mctppcc%d", context.inbox_index);
+	ndev = alloc_netdev(sizeof(*mctp_pcc_ndev), name, NET_NAME_PREDICTABLE,
+			    mctp_pcc_setup);
+	if (!ndev)
+		return -ENOMEM;
+
+	mctp_pcc_ndev = netdev_priv(ndev);
+
+	mctp_pcc_ndev->inbox.index = context.inbox_index;
+	mctp_pcc_ndev->inbox.client.dev = dev;
+	mctp_pcc_ndev->outbox.index = context.outbox_index;
+	mctp_pcc_ndev->outbox.client.dev = dev;
+
+	mctp_pcc_ndev->outbox.client.tx_prepare = mctp_pcc_tx_prepare;
+	mctp_pcc_ndev->outbox.client.tx_done = mctp_pcc_tx_done;
+	mctp_pcc_ndev->acpi_device = acpi_dev;
+	mctp_pcc_ndev->ndev = ndev;
+	acpi_dev->driver_data = mctp_pcc_ndev;
+
+	rc = initialize_mtu(ndev);
+	if (rc)
+		goto free_netdev;
+
+	rc = mctp_register_netdev(ndev, NULL, MCTP_PHYS_BINDING_PCC);
+	if (rc)
+		goto free_netdev;
+
+	return devm_add_action_or_reset(dev, mctp_cleanup_netdev, ndev);
+free_netdev:
+	free_netdev(ndev);
+	return rc;
+}
+
+static const struct acpi_device_id mctp_pcc_device_ids[] = {
+	{ "DMT0001" },
+	{}
+};
+
+static struct acpi_driver mctp_pcc_driver = {
+	.name = "mctp_pcc",
+	.class = "Unknown",
+	.ids = mctp_pcc_device_ids,
+	.ops = {
+		.add = mctp_pcc_driver_add,
+	},
+};
+
+module_acpi_driver(mctp_pcc_driver);
+
+MODULE_DEVICE_TABLE(acpi, mctp_pcc_device_ids);
+
+MODULE_DESCRIPTION("MCTP PCC ACPI device");
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Adam Young <admiyo@os.amperecomputing.com>");
-- 
2.43.0

Re: [net-next v41] mctp pcc: Implement MCTP over PCC Transport

[Jeremy Kerr]

Hi Adam,

You keep sending out versions of this patch which have not accommodated
items of previous feedback. It is not clear if this is intentional, but
if you do disagree with some request for change, then respond to that
request, rather than silently ignoring it.

Alternatively, if it's not clear what the request is, then please
respond to that thread and ask for clarification.

Seeing changes without previous change requests either implemented or
rejected through discussion is making the review process pretty
frustrating, and is a primary reason why this has taken 40 versions so
far.

> Changes from Previous version
> 
> Remove check for skb_is_nonlinear(skb) as it is done in skb_linearize(skb)
> Removed comment about BE support
> Spacing changes after gotos

This change is good, but hasn't been applied to the part of the driver I
had commented on.

> +static void mctp_pcc_client_rx_callback(struct mbox_client *cl, void *mssg)
> +{
> +       struct acpi_pcct_ext_pcc_shared_memory pcc_header;
> +       struct mctp_pcc_ndev *mctp_pcc_ndev;
> +       struct mctp_pcc_mailbox *inbox;
> +       struct mctp_skb_cb *cb;
> +       struct sk_buff *skb;
> +       u32 header_length;
> +       int size;
> +
> +       mctp_pcc_ndev = container_of(cl, struct mctp_pcc_ndev, inbox.client);
> +       inbox = &mctp_pcc_ndev->inbox;
> +       memcpy_fromio(&pcc_header, inbox->chan->shmem, sizeof(pcc_header));
> +
> +       // The message must at least have the PCC command indicating it is an MCTP
> +       // message followed by the MCTP header, or we have a malformed message.
> +       // This may be run on big endian system, but the data in the buffer is
> +       // explicitly little endian.
> +       header_length = le32_to_cpu(pcc_header.length);
> +
> +       if (header_length < sizeof(pcc_header.command) + sizeof(struct mctp_hdr))
> +               goto error;
> +
> +       // If the reported size is larger than the shared memory minus headers,
> +       // something is wrong and treat the buffer as corrupted data.
> +       if (header_length > inbox->chan->shmem_size - PCC_EXTRA_LEN)
> +               goto error;
> +
> +       if (memcmp(&pcc_header.command, MCTP_SIGNATURE, MCTP_SIGNATURE_LENGTH) != 0)
> +               goto error;
> +
> +       size = header_length + PCC_EXTRA_LEN;
> +       skb = netdev_alloc_skb(mctp_pcc_ndev->ndev, size);
> +       if (!skb)
> +               goto error;

As I mentioned on v40, space after this.

> +       skb_put(skb, size);
> +       skb->protocol = htons(ETH_P_MCTP);
> +       memcpy_fromio(skb->data, inbox->chan->shmem, size);
> +       dev_dstats_rx_add(mctp_pcc_ndev->ndev, size);
> +       skb_pull(skb, sizeof(pcc_header));
> +       skb_reset_mac_header(skb);
> +       skb_reset_network_header(skb);
> +       cb = __mctp_cb(skb);
> +       cb->halen = 0;
> +       netif_rx(skb);
> +       return;

And one after this.

> +error:
> +       dev_dstats_rx_dropped(mctp_pcc_ndev->ndev);
> +}
> +
> +static netdev_tx_t mctp_pcc_tx(struct sk_buff *skb, struct net_device *ndev)
> +{
> +       struct acpi_pcct_ext_pcc_shared_memory *pcc_header;
> +       struct mctp_pcc_ndev *mpnd = netdev_priv(ndev);
> +       int len = skb->len;
> +
> +       /* Consolidated a fragmented packet into contiguous memory */
> +       if (skb_linearize(skb))
> +               goto error;

You have removed the unnecessary is_linear check, but the entire
linearize is unnecessary. From v39:

  skb_linearize() already has the skb_is_nonlinear() check.
  
  However, you don't need to call skb_linearize() anyway, as that will
  happen for you in validate_xmit_skb(), since the driver does not
  advertise support for nonlinear skbs.

> +
> +       if (skb_cow_head(skb, sizeof(*pcc_header)))
> +               goto error;

... and, if you're going for spacing updates, may as well do one here
too

> +       pcc_header = skb_push(skb, sizeof(*pcc_header));
> +       pcc_header->signature = PCC_SIGNATURE | mpnd->outbox.index;
> +       pcc_header->flags = PCC_CMD_COMPLETION_NOTIFY;
> +       memcpy(&pcc_header->command, MCTP_SIGNATURE, MCTP_SIGNATURE_LENGTH);
> +       pcc_header->length = len + MCTP_SIGNATURE_LENGTH;
> +
> +       if (mbox_send_message(mpnd->outbox.chan->mchan, skb) < 0) {
> +               // Remove the header in case it gets sent again
> +               skb_pull(skb, sizeof(*pcc_header));
> +               netif_stop_queue(ndev);
> +               return NETDEV_TX_BUSY;
> +       }
> +
> +       return NETDEV_TX_OK;

And here.

> +error:
> +       dev_dstats_tx_dropped(ndev);
> +       kfree_skb(skb);
> +       return NETDEV_TX_OK;
> +}
> +
> +static void mctp_pcc_tx_prepare(struct mbox_client *cl, void *mssg)
> +{
> +       struct mctp_pcc_ndev *mctp_pcc_ndev;
> +       struct mctp_pcc_mailbox *outbox;
> +       struct sk_buff *skb = mssg;
> +
> +       mctp_pcc_ndev = container_of(cl, struct mctp_pcc_ndev, outbox.client);
> +       outbox = &mctp_pcc_ndev->outbox;
> +
> +       /* The PCC Mailbox typically does not make use of the mssg pointer
> +        * The mctp-over pcc driver is the only client that uses it.
> +        * This value should always be non-null; it is possible
> +        * that a change in the Mailbox level will break that assumption.
> +        */
> +       if (!skb) {
> +               netdev_warn_once(mctp_pcc_ndev->ndev,
> +                                "%s called with null message.\n", __func__);
> +               return;
> +       }
> +
> +       if (skb->len > outbox->chan->shmem_size) {
> +               dev_dstats_tx_dropped(mctp_pcc_ndev->ndev);
> +               return;
> +       }
> +       memcpy_toio(outbox->chan->shmem, skb->data, skb->len);
> +
> +       /*
> +        * This packet could still be dropped in the PCC layer,
> +        * But the only place that could deal with that is mctp_pcc_tx_done.
> +        * That is called from the Hard-IRQ handler, and it is not safe to
> +        * call stats functions from HARD-IRQ context.
> +        */

Your other option, if you prefer, is to do an irq-safe stats update.
This is essentially open-coding dev_dstats_tx_add(), but using the
irqsave/irqrestore variants of begin/end. Something like:

	struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
	unsigned long flags;

	flags = u64_stats_update_begin_irqsave(&dstats->syncp);
	u64_stats_inc(&dstats->tx_packets);
	u64_stats_add(&dstats->tx_bytes, skb->len);
	u64_stats_update_end_irqrestore(&dstats->syncp, flags);

(but your current approach is okay too)

Cheers,


Jeremy

Re: [net-next v41] mctp pcc: Implement MCTP over PCC Transport

[Adam Young]

On 5/10/26 22:19, Jeremy Kerr wrote:
> Hi Adam,
>
> You keep sending out versions of this patch which have not accommodated
> items of previous feedback. It is not clear if this is intentional, but
> if you do disagree with some request for change, then respond to that
> request, rather than silently ignoring it.
>
> Alternatively, if it's not clear what the request is, then please
> respond to that thread and ask for clarification.
>
> Seeing changes without previous change requests either implemented or
> rejected through discussion is making the review process pretty
> frustrating, and is a primary reason why this has taken 40 versions so
> far.

Sorry, I should have been more explicit here.  I am not certain what is 
going to happen with fragmentation, so I want to be protected against 
future changes.

The check in validate_xmit_skb() is good, as it protects against the 
current set up. So my option was to put a comment in here and hope both 
changes happened together, or to just try and get this portion of the 
driver solid against the change.

And I thought that was what you were suggesting in the comment. The 
original comment sounded more like an "here is an optimization" instead 
of "this is important enough to kick back"

As for spacing, I get that there is a style, but it really should be 
encoded in checkstyle.sh or something and automated.  My own tendency is 
to put way too many spaces in to chunk things together, and I end up 
going over-draconian on stripping them out to try and meet the expected 
layout.

So, no, I am not intentionally skipping things.  I am really, really 
trying to catch all the nits and get this driver into an acceptable 
state.  I really appreciate the effort you have put in to review, as it 
would be dead in the water without your feedback.


>
>> Changes from Previous version
>>
>> Remove check for skb_is_nonlinear(skb) as it is done in skb_linearize(skb)
>> Removed comment about BE support
>> Spacing changes after gotos
> This change is good, but hasn't been applied to the part of the driver I
> had commented on.
>
>> +static void mctp_pcc_client_rx_callback(struct mbox_client *cl, void *mssg)
>> +{
>> +       struct acpi_pcct_ext_pcc_shared_memory pcc_header;
>> +       struct mctp_pcc_ndev *mctp_pcc_ndev;
>> +       struct mctp_pcc_mailbox *inbox;
>> +       struct mctp_skb_cb *cb;
>> +       struct sk_buff *skb;
>> +       u32 header_length;
>> +       int size;
>> +
>> +       mctp_pcc_ndev = container_of(cl, struct mctp_pcc_ndev, inbox.client);
>> +       inbox = &mctp_pcc_ndev->inbox;
>> +       memcpy_fromio(&pcc_header, inbox->chan->shmem, sizeof(pcc_header));
>> +
>> +       // The message must at least have the PCC command indicating it is an MCTP
>> +       // message followed by the MCTP header, or we have a malformed message.
>> +       // This may be run on big endian system, but the data in the buffer is
>> +       // explicitly little endian.
>> +       header_length = le32_to_cpu(pcc_header.length);
>> +
>> +       if (header_length < sizeof(pcc_header.command) + sizeof(struct mctp_hdr))
>> +               goto error;
>> +
>> +       // If the reported size is larger than the shared memory minus headers,
>> +       // something is wrong and treat the buffer as corrupted data.
>> +       if (header_length > inbox->chan->shmem_size - PCC_EXTRA_LEN)
>> +               goto error;
>> +
>> +       if (memcmp(&pcc_header.command, MCTP_SIGNATURE, MCTP_SIGNATURE_LENGTH) != 0)
>> +               goto error;
>> +
>> +       size = header_length + PCC_EXTRA_LEN;
>> +       skb = netdev_alloc_skb(mctp_pcc_ndev->ndev, size);
>> +       if (!skb)
>> +               goto error;
> As I mentioned on v40, space after this.
Sorry, I have been reformatting a bunch.  Added it in and then removed it.
>
>> +       skb_put(skb, size);
>> +       skb->protocol = htons(ETH_P_MCTP);
>> +       memcpy_fromio(skb->data, inbox->chan->shmem, size);
>> +       dev_dstats_rx_add(mctp_pcc_ndev->ndev, size);
>> +       skb_pull(skb, sizeof(pcc_header));
>> +       skb_reset_mac_header(skb);
>> +       skb_reset_network_header(skb);
>> +       cb = __mctp_cb(skb);
>> +       cb->halen = 0;
>> +       netif_rx(skb);
>> +       return;
> And one after this.
>
>> +error:
>> +       dev_dstats_rx_dropped(mctp_pcc_ndev->ndev);
>> +}
>> +
>> +static netdev_tx_t mctp_pcc_tx(struct sk_buff *skb, struct net_device *ndev)
>> +{
>> +       struct acpi_pcct_ext_pcc_shared_memory *pcc_header;
>> +       struct mctp_pcc_ndev *mpnd = netdev_priv(ndev);
>> +       int len = skb->len;
>> +
>> +       /* Consolidated a fragmented packet into contiguous memory */
>> +       if (skb_linearize(skb))
>> +               goto error;
> You have removed the unnecessary is_linear check, but the entire
> linearize is unnecessary. From v39:
>
>    skb_linearize() already has the skb_is_nonlinear() check.
>    
>    However, you don't need to call skb_linearize() anyway, as that will
>    happen for you in validate_xmit_skb(), since the driver does not
>    advertise support for nonlinear skbs.
>
>> +
>> +       if (skb_cow_head(skb, sizeof(*pcc_header)))
>> +               goto error;
> ... and, if you're going for spacing updates, may as well do one here
> too
>
>> +       pcc_header = skb_push(skb, sizeof(*pcc_header));
>> +       pcc_header->signature = PCC_SIGNATURE | mpnd->outbox.index;
>> +       pcc_header->flags = PCC_CMD_COMPLETION_NOTIFY;
>> +       memcpy(&pcc_header->command, MCTP_SIGNATURE, MCTP_SIGNATURE_LENGTH);
>> +       pcc_header->length = len + MCTP_SIGNATURE_LENGTH;
>> +
>> +       if (mbox_send_message(mpnd->outbox.chan->mchan, skb) < 0) {
>> +               // Remove the header in case it gets sent again
>> +               skb_pull(skb, sizeof(*pcc_header));
>> +               netif_stop_queue(ndev);
>> +               return NETDEV_TX_BUSY;
>> +       }
>> +
>> +       return NETDEV_TX_OK;
> And here.
>
>> +error:
>> +       dev_dstats_tx_dropped(ndev);
>> +       kfree_skb(skb);
>> +       return NETDEV_TX_OK;
>> +}
>> +
>> +static void mctp_pcc_tx_prepare(struct mbox_client *cl, void *mssg)
>> +{
>> +       struct mctp_pcc_ndev *mctp_pcc_ndev;
>> +       struct mctp_pcc_mailbox *outbox;
>> +       struct sk_buff *skb = mssg;
>> +
>> +       mctp_pcc_ndev = container_of(cl, struct mctp_pcc_ndev, outbox.client);
>> +       outbox = &mctp_pcc_ndev->outbox;
>> +
>> +       /* The PCC Mailbox typically does not make use of the mssg pointer
>> +        * The mctp-over pcc driver is the only client that uses it.
>> +        * This value should always be non-null; it is possible
>> +        * that a change in the Mailbox level will break that assumption.
>> +        */
>> +       if (!skb) {
>> +               netdev_warn_once(mctp_pcc_ndev->ndev,
>> +                                "%s called with null message.\n", __func__);
>> +               return;
>> +       }
>> +
>> +       if (skb->len > outbox->chan->shmem_size) {
>> +               dev_dstats_tx_dropped(mctp_pcc_ndev->ndev);
>> +               return;
>> +       }
>> +       memcpy_toio(outbox->chan->shmem, skb->data, skb->len);
>> +
>> +       /*
>> +        * This packet could still be dropped in the PCC layer,
>> +        * But the only place that could deal with that is mctp_pcc_tx_done.
>> +        * That is called from the Hard-IRQ handler, and it is not safe to
>> +        * call stats functions from HARD-IRQ context.
>> +        */
> Your other option, if you prefer, is to do an irq-safe stats update.
> This is essentially open-coding dev_dstats_tx_add(), but using the
> irqsave/irqrestore variants of begin/end. Something like:
>
> 	struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
> 	unsigned long flags;
>
> 	flags = u64_stats_update_begin_irqsave(&dstats->syncp);
> 	u64_stats_inc(&dstats->tx_packets);
> 	u64_stats_add(&dstats->tx_bytes, skb->len);
> 	u64_stats_update_end_irqrestore(&dstats->syncp, flags);
>
> (but your current approach is okay too

Thanks for the approach.  I will possibly incorporate this in a future 
patch.  I have another approach in mind I want to try first.


>
> Cheers,
>
>
> Jeremy