Claude review: gpu: nova-core: gsp: clarify invariant on command queue

Claude review: gpu: nova-core: gsp: clarify invariant on command queue

[Claude Code Review Bot]

Patch Review

> +        // Extract area for the command itself. The GSP message header and the command header
> +        // together are guaranteed to fit entirely into a single page, so it's ok to only look
> +        // at `dst.contents.0` here.

Comment-only change adding useful context about why `from_bytes_mut_prefix` only uses the first content slice. No issues.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: clarify invariant on command queue

[Claude Code Review Bot]

Patch Review

Adds a comment explaining why only `dst.contents.0` is used for the command header extraction:

```rust
// The GSP message header and the command header
// together are guaranteed to fit entirely into a single page, so it's ok to only look
// at `dst.contents.0` here.
```

This is a useful clarification. The guarantee holds because `GspMsgElement` + any `Command` struct header is well under 4KB (one GSP page). The comment helps readers understand why the wrap-around second slice isn't needed for the header portion.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: clarify invariant on command queue

[Claude Code Review Bot]

Patch Review

Adds a comment explaining why only `dst.contents.0` is used for the command header extraction:

```rust
// Extract area for the command itself. The GSP message header and the command header
// together are guaranteed to fit entirely into a single page, so it's ok to only look
// at `dst.contents.0` here.
```

Good clarification. Since `GspMsgElement` is 64 bytes and command headers are small, they'll always fit within the first contiguous slice of the circular buffer.

---
Generated by Claude Code Patch Reviewer

[PATCH v5 0/9] gpu: nova-core: gsp: add continuation record support

[Eliot Courtney]

GSP commands over 16 pages need to be sent using "continuation records"
which essentially means splitting the payload over multiple commands.

This series adds a command type `ContinuationRecord` which just writes
its header and whatever payload it is given. It also adds types
`SplitState` and `SplitCommand` which support splitting a large RPC
into smaller ones while transparently letting regular sized RPCs be sent
without extra copies.

The send pathway uses `SplitState` to send all commands, but if
the command fits into 16 pages, it still writes directly into the
command queue. If it is larger than 16 pages and needs continuation
records, it writes into a staging buffer, so there is one copy.

Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
Changes in v5:
- Moved continuation module under cmdq.
- Made timeout into a constant.
- Renamed command_size to size_in_bytes and made it a default method
  on CommandToGsp
- Link to v4: https://lore.kernel.org/r/20260302-cmdq-continuation-v4-0-c011f15aad58@nvidia.com

Changes in v4:
- Removed Tested-by Zhi on the ContinuationRecord patch, since it
  changed substantively.
- Incorporated Alex's design for SplitState
- Moved the continuation record types into their own module.
- Tightened visibility on continuation record types.
- Used non-zero sized command in tests.
- Link to v3: https://lore.kernel.org/r/20260226-cmdq-continuation-v3-0-572ab9916766@nvidia.com

Changes in v3:
- Rebased onto drm-rust-next
- Added some doc comments
- Reworded command_size doc comment
- Link to v2: https://lore.kernel.org/r/20260219-cmdq-continuation-v2-0-2e8b7615536f@nvidia.com

Changes in v2:
- Added doccoments
- Renamed driver_bytes_available_to_write to driver_write_area_size
- allocate_command_with_timeout merged allocate_command with timeout parameter
- Replaced hardcoded GSP_PAGE_SIZE * 16 with bindings
- Changed oversized command error from EIO to EMSGSIZE
- Added EMSGSIZE to kernel/error.rs
- Split WrappingCommand functionality into SplitState + SplitCommand enum
- Made max_size a const (MAX_CMD_SIZE)
- Removed send_continuation_record + added comment for type inference
- send_single_command now consumes the command
- Extracted command_size + used in SplitState 
- Link to v1: https://lore.kernel.org/r/20260212-cmdq-continuation-v1-0-73079ded55e6@nvidia.com

---
Eliot Courtney (9):
      gpu: nova-core: gsp: sort `MsgFunction` variants alphabetically
      gpu: nova-core: gsp: add mechanism to wait for space on command queue
      rust: add EMSGSIZE error code
      gpu: nova-core: gsp: add checking oversized commands
      gpu: nova-core: gsp: clarify invariant on command queue
      gpu: nova-core: gsp: unconditionally call variable payload handling
      gpu: nova-core: gsp: add `size_in_bytes` helper to `CommandToGsp`
      gpu: nova-core: gsp: support large RPCs via continuation record
      gpu: nova-core: gsp: add tests for continuation records

 drivers/gpu/nova-core/gsp/cmdq.rs                 | 119 +++++++--
 drivers/gpu/nova-core/gsp/cmdq/continuation.rs    | 301 ++++++++++++++++++++++
 drivers/gpu/nova-core/gsp/fw.rs                   |  75 +++---
 drivers/gpu/nova-core/gsp/fw/r570_144/bindings.rs |   1 +
 rust/kernel/error.rs                              |   1 +
 5 files changed, 439 insertions(+), 58 deletions(-)
---
base-commit: 4a49fe23e357b48845e31fe9c28a802c05458198
change-id: 20260203-cmdq-continuation-b99f3d5966c3

Best regards,
-- 
Eliot Courtney <ecourtney@nvidia.com>

[PATCH v5 1/9] gpu: nova-core: gsp: sort `MsgFunction` variants alphabetically

[Eliot Courtney]

There is no particular order required here and keeping them alphabetical
will help preventing future mistakes.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/fw.rs | 67 +++++++++++++++++++++--------------------
 1 file changed, 35 insertions(+), 32 deletions(-)

diff --git a/drivers/gpu/nova-core/gsp/fw.rs b/drivers/gpu/nova-core/gsp/fw.rs
index f1797e1f0d9d..4b998485360b 100644
--- a/drivers/gpu/nova-core/gsp/fw.rs
+++ b/drivers/gpu/nova-core/gsp/fw.rs
@@ -191,34 +191,34 @@ pub(crate) fn new(gsp_firmware: &GspFirmware, fb_layout: &FbLayout) -> Self {
 #[repr(u32)]
 pub(crate) enum MsgFunction {
     // Common function codes
-    Nop = bindings::NV_VGPU_MSG_FUNCTION_NOP,
-    SetGuestSystemInfo = bindings::NV_VGPU_MSG_FUNCTION_SET_GUEST_SYSTEM_INFO,
-    AllocRoot = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_ROOT,
+    AllocChannelDma = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CHANNEL_DMA,
+    AllocCtxDma = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CTX_DMA,
     AllocDevice = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_DEVICE,
     AllocMemory = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_MEMORY,
-    AllocCtxDma = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CTX_DMA,
-    AllocChannelDma = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CHANNEL_DMA,
-    MapMemory = bindings::NV_VGPU_MSG_FUNCTION_MAP_MEMORY,
-    BindCtxDma = bindings::NV_VGPU_MSG_FUNCTION_BIND_CTX_DMA,
     AllocObject = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_OBJECT,
+    AllocRoot = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_ROOT,
+    BindCtxDma = bindings::NV_VGPU_MSG_FUNCTION_BIND_CTX_DMA,
     Free = bindings::NV_VGPU_MSG_FUNCTION_FREE,
-    Log = bindings::NV_VGPU_MSG_FUNCTION_LOG,
     GetGspStaticInfo = bindings::NV_VGPU_MSG_FUNCTION_GET_GSP_STATIC_INFO,
-    SetRegistry = bindings::NV_VGPU_MSG_FUNCTION_SET_REGISTRY,
-    GspSetSystemInfo = bindings::NV_VGPU_MSG_FUNCTION_GSP_SET_SYSTEM_INFO,
+    GetStaticInfo = bindings::NV_VGPU_MSG_FUNCTION_GET_STATIC_INFO,
     GspInitPostObjGpu = bindings::NV_VGPU_MSG_FUNCTION_GSP_INIT_POST_OBJGPU,
     GspRmControl = bindings::NV_VGPU_MSG_FUNCTION_GSP_RM_CONTROL,
-    GetStaticInfo = bindings::NV_VGPU_MSG_FUNCTION_GET_STATIC_INFO,
+    GspSetSystemInfo = bindings::NV_VGPU_MSG_FUNCTION_GSP_SET_SYSTEM_INFO,
+    Log = bindings::NV_VGPU_MSG_FUNCTION_LOG,
+    MapMemory = bindings::NV_VGPU_MSG_FUNCTION_MAP_MEMORY,
+    Nop = bindings::NV_VGPU_MSG_FUNCTION_NOP,
+    SetGuestSystemInfo = bindings::NV_VGPU_MSG_FUNCTION_SET_GUEST_SYSTEM_INFO,
+    SetRegistry = bindings::NV_VGPU_MSG_FUNCTION_SET_REGISTRY,
 
     // Event codes
     GspInitDone = bindings::NV_VGPU_MSG_EVENT_GSP_INIT_DONE,
+    GspLockdownNotice = bindings::NV_VGPU_MSG_EVENT_GSP_LOCKDOWN_NOTICE,
+    GspPostNoCat = bindings::NV_VGPU_MSG_EVENT_GSP_POST_NOCAT_RECORD,
     GspRunCpuSequencer = bindings::NV_VGPU_MSG_EVENT_GSP_RUN_CPU_SEQUENCER,
-    PostEvent = bindings::NV_VGPU_MSG_EVENT_POST_EVENT,
-    RcTriggered = bindings::NV_VGPU_MSG_EVENT_RC_TRIGGERED,
     MmuFaultQueued = bindings::NV_VGPU_MSG_EVENT_MMU_FAULT_QUEUED,
     OsErrorLog = bindings::NV_VGPU_MSG_EVENT_OS_ERROR_LOG,
-    GspPostNoCat = bindings::NV_VGPU_MSG_EVENT_GSP_POST_NOCAT_RECORD,
-    GspLockdownNotice = bindings::NV_VGPU_MSG_EVENT_GSP_LOCKDOWN_NOTICE,
+    PostEvent = bindings::NV_VGPU_MSG_EVENT_POST_EVENT,
+    RcTriggered = bindings::NV_VGPU_MSG_EVENT_RC_TRIGGERED,
     UcodeLibOsPrint = bindings::NV_VGPU_MSG_EVENT_UCODE_LIBOS_PRINT,
 }
 
@@ -227,38 +227,41 @@ impl TryFrom<u32> for MsgFunction {
 
     fn try_from(value: u32) -> Result<MsgFunction> {
         match value {
-            bindings::NV_VGPU_MSG_FUNCTION_NOP => Ok(MsgFunction::Nop),
-            bindings::NV_VGPU_MSG_FUNCTION_SET_GUEST_SYSTEM_INFO => {
-                Ok(MsgFunction::SetGuestSystemInfo)
-            }
-            bindings::NV_VGPU_MSG_FUNCTION_ALLOC_ROOT => Ok(MsgFunction::AllocRoot),
+            // Common function codes
+            bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CHANNEL_DMA => Ok(MsgFunction::AllocChannelDma),
+            bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CTX_DMA => Ok(MsgFunction::AllocCtxDma),
             bindings::NV_VGPU_MSG_FUNCTION_ALLOC_DEVICE => Ok(MsgFunction::AllocDevice),
             bindings::NV_VGPU_MSG_FUNCTION_ALLOC_MEMORY => Ok(MsgFunction::AllocMemory),
-            bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CTX_DMA => Ok(MsgFunction::AllocCtxDma),
-            bindings::NV_VGPU_MSG_FUNCTION_ALLOC_CHANNEL_DMA => Ok(MsgFunction::AllocChannelDma),
-            bindings::NV_VGPU_MSG_FUNCTION_MAP_MEMORY => Ok(MsgFunction::MapMemory),
-            bindings::NV_VGPU_MSG_FUNCTION_BIND_CTX_DMA => Ok(MsgFunction::BindCtxDma),
             bindings::NV_VGPU_MSG_FUNCTION_ALLOC_OBJECT => Ok(MsgFunction::AllocObject),
+            bindings::NV_VGPU_MSG_FUNCTION_ALLOC_ROOT => Ok(MsgFunction::AllocRoot),
+            bindings::NV_VGPU_MSG_FUNCTION_BIND_CTX_DMA => Ok(MsgFunction::BindCtxDma),
             bindings::NV_VGPU_MSG_FUNCTION_FREE => Ok(MsgFunction::Free),
-            bindings::NV_VGPU_MSG_FUNCTION_LOG => Ok(MsgFunction::Log),
             bindings::NV_VGPU_MSG_FUNCTION_GET_GSP_STATIC_INFO => Ok(MsgFunction::GetGspStaticInfo),
-            bindings::NV_VGPU_MSG_FUNCTION_SET_REGISTRY => Ok(MsgFunction::SetRegistry),
-            bindings::NV_VGPU_MSG_FUNCTION_GSP_SET_SYSTEM_INFO => Ok(MsgFunction::GspSetSystemInfo),
+            bindings::NV_VGPU_MSG_FUNCTION_GET_STATIC_INFO => Ok(MsgFunction::GetStaticInfo),
             bindings::NV_VGPU_MSG_FUNCTION_GSP_INIT_POST_OBJGPU => {
                 Ok(MsgFunction::GspInitPostObjGpu)
             }
             bindings::NV_VGPU_MSG_FUNCTION_GSP_RM_CONTROL => Ok(MsgFunction::GspRmControl),
-            bindings::NV_VGPU_MSG_FUNCTION_GET_STATIC_INFO => Ok(MsgFunction::GetStaticInfo),
+            bindings::NV_VGPU_MSG_FUNCTION_GSP_SET_SYSTEM_INFO => Ok(MsgFunction::GspSetSystemInfo),
+            bindings::NV_VGPU_MSG_FUNCTION_LOG => Ok(MsgFunction::Log),
+            bindings::NV_VGPU_MSG_FUNCTION_MAP_MEMORY => Ok(MsgFunction::MapMemory),
+            bindings::NV_VGPU_MSG_FUNCTION_NOP => Ok(MsgFunction::Nop),
+            bindings::NV_VGPU_MSG_FUNCTION_SET_GUEST_SYSTEM_INFO => {
+                Ok(MsgFunction::SetGuestSystemInfo)
+            }
+            bindings::NV_VGPU_MSG_FUNCTION_SET_REGISTRY => Ok(MsgFunction::SetRegistry),
+
+            // Event codes
             bindings::NV_VGPU_MSG_EVENT_GSP_INIT_DONE => Ok(MsgFunction::GspInitDone),
+            bindings::NV_VGPU_MSG_EVENT_GSP_LOCKDOWN_NOTICE => Ok(MsgFunction::GspLockdownNotice),
+            bindings::NV_VGPU_MSG_EVENT_GSP_POST_NOCAT_RECORD => Ok(MsgFunction::GspPostNoCat),
             bindings::NV_VGPU_MSG_EVENT_GSP_RUN_CPU_SEQUENCER => {
                 Ok(MsgFunction::GspRunCpuSequencer)
             }
-            bindings::NV_VGPU_MSG_EVENT_POST_EVENT => Ok(MsgFunction::PostEvent),
-            bindings::NV_VGPU_MSG_EVENT_RC_TRIGGERED => Ok(MsgFunction::RcTriggered),
             bindings::NV_VGPU_MSG_EVENT_MMU_FAULT_QUEUED => Ok(MsgFunction::MmuFaultQueued),
             bindings::NV_VGPU_MSG_EVENT_OS_ERROR_LOG => Ok(MsgFunction::OsErrorLog),
-            bindings::NV_VGPU_MSG_EVENT_GSP_POST_NOCAT_RECORD => Ok(MsgFunction::GspPostNoCat),
-            bindings::NV_VGPU_MSG_EVENT_GSP_LOCKDOWN_NOTICE => Ok(MsgFunction::GspLockdownNotice),
+            bindings::NV_VGPU_MSG_EVENT_POST_EVENT => Ok(MsgFunction::PostEvent),
+            bindings::NV_VGPU_MSG_EVENT_RC_TRIGGERED => Ok(MsgFunction::RcTriggered),
             bindings::NV_VGPU_MSG_EVENT_UCODE_LIBOS_PRINT => Ok(MsgFunction::UcodeLibOsPrint),
             _ => Err(EINVAL),
         }

-- 
2.53.0

[PATCH v5 2/9] gpu: nova-core: gsp: add mechanism to wait for space on command queue

[Eliot Courtney]

Add a timeout to `allocate_command` which waits for space on the GSP
command queue. It uses a similar timeout to nouveau.

This lets `send_command` wait for space to free up in the command queue.
This is required to support continuation records which can fill up the
queue.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq.rs | 42 ++++++++++++++++++++++++++++-----------
 1 file changed, 30 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
index 87dbbd6d1be9..5cb845bae123 100644
--- a/drivers/gpu/nova-core/gsp/cmdq.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq.rs
@@ -250,6 +250,19 @@ fn new(dev: &device::Device<device::Bound>) -> Result<Self> {
         }
     }
 
+    /// Returns the size of the region of the CPU message queue that the driver is currently allowed
+    /// to write to, in bytes.
+    fn driver_write_area_size(&self) -> usize {
+        let tx = self.cpu_write_ptr();
+        let rx = self.gsp_read_ptr();
+
+        // `rx` and `tx` are both in `0..MSGQ_NUM_PAGES` per the invariants of `gsp_read_ptr` and
+        // `cpu_write_ptr`. The minimum value case is where `rx == 0` and `tx == MSGQ_NUM_PAGES -
+        // 1`, which gives `0 + MSGQ_NUM_PAGES - (MSGQ_NUM_PAGES - 1) - 1 == 0`.
+        let slots = (rx + MSGQ_NUM_PAGES - tx - 1) % MSGQ_NUM_PAGES;
+        num::u32_as_usize(slots) * GSP_PAGE_SIZE
+    }
+
     /// Returns the region of the GSP message queue that the driver is currently allowed to read
     /// from.
     ///
@@ -281,15 +294,22 @@ fn new(dev: &device::Device<device::Bound>) -> Result<Self> {
     }
 
     /// Allocates a region on the command queue that is large enough to send a command of `size`
-    /// bytes.
+    /// bytes, waiting for space to become available based on the provided timeout.
     ///
     /// This returns a [`GspCommand`] ready to be written to by the caller.
     ///
     /// # Errors
     ///
-    /// - `EAGAIN` if the driver area is too small to hold the requested command.
+    /// - `ETIMEDOUT` if space does not become available within the timeout.
     /// - `EIO` if the command header is not properly aligned.
-    fn allocate_command(&mut self, size: usize) -> Result<GspCommand<'_>> {
+    fn allocate_command(&mut self, size: usize, timeout: Delta) -> Result<GspCommand<'_>> {
+        read_poll_timeout(
+            || Ok(self.driver_write_area_size()),
+            |available_bytes| *available_bytes >= size_of::<GspMsgElement>() + size,
+            Delta::ZERO,
+            timeout,
+        )?;
+
         // Get the current writable area as an array of bytes.
         let (slice_1, slice_2) = {
             let (slice_1, slice_2) = self.driver_write_area();
@@ -298,13 +318,6 @@ fn allocate_command(&mut self, size: usize) -> Result<GspCommand<'_>> {
             (slice_1.as_flattened_mut(), slice_2.as_flattened_mut())
         };
 
-        // If the GSP is still processing previous messages the shared region
-        // may be full in which case we will have to retry once the GSP has
-        // processed the existing commands.
-        if size_of::<GspMsgElement>() + size > slice_1.len() + slice_2.len() {
-            return Err(EAGAIN);
-        }
-
         // Extract area for the `GspMsgElement`.
         let (header, slice_1) = GspMsgElement::from_bytes_mut_prefix(slice_1).ok_or(EIO)?;
 
@@ -462,6 +475,9 @@ impl Cmdq {
     /// Number of page table entries for the GSP shared region.
     pub(crate) const NUM_PTES: usize = size_of::<GspMem>() >> GSP_PAGE_SHIFT;
 
+    /// Timeout for waiting for space on the command queue.
+    const ALLOCATE_TIMEOUT: Delta = Delta::from_secs(1);
+
     /// Creates a new command queue for `dev`.
     pub(crate) fn new(dev: &device::Device<device::Bound>) -> Result<Cmdq> {
         let gsp_mem = DmaGspMem::new(dev)?;
@@ -497,7 +513,7 @@ fn notify_gsp(bar: &Bar0) {
     ///
     /// # Errors
     ///
-    /// - `EAGAIN` if there was not enough space in the command queue to send the command.
+    /// - `ETIMEDOUT` if space does not become available within the timeout.
     /// - `EIO` if the variable payload requested by the command has not been entirely
     ///   written to by its [`CommandToGsp::init_variable_payload`] method.
     ///
@@ -509,7 +525,9 @@ pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
         Error: From<M::InitError>,
     {
         let command_size = size_of::<M::Command>() + command.variable_payload_len();
-        let dst = self.gsp_mem.allocate_command(command_size)?;
+        let dst = self
+            .gsp_mem
+            .allocate_command(command_size, Self::ALLOCATE_TIMEOUT)?;
 
         // Extract area for the command itself.
         let (cmd, payload_1) = M::Command::from_bytes_mut_prefix(dst.contents.0).ok_or(EIO)?;

-- 
2.53.0

[PATCH v5 3/9] rust: add EMSGSIZE error code

[Eliot Courtney]

Add the EMSGSIZE error code, which indicates that a message is too
long.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Reviewed-by: Gary Guo <gary@garyguo.net>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 rust/kernel/error.rs | 1 +
 1 file changed, 1 insertion(+)

diff --git a/rust/kernel/error.rs b/rust/kernel/error.rs
index 258b12afdcba..10fcf1f0404d 100644
--- a/rust/kernel/error.rs
+++ b/rust/kernel/error.rs
@@ -67,6 +67,7 @@ macro_rules! declare_err {
     declare_err!(EDOM, "Math argument out of domain of func.");
     declare_err!(ERANGE, "Math result not representable.");
     declare_err!(EOVERFLOW, "Value too large for defined data type.");
+    declare_err!(EMSGSIZE, "Message too long.");
     declare_err!(ETIMEDOUT, "Connection timed out.");
     declare_err!(ERESTARTSYS, "Restart the system call.");
     declare_err!(ERESTARTNOINTR, "System call was interrupted by a signal and will be restarted.");

-- 
2.53.0

[PATCH v5 4/9] gpu: nova-core: gsp: add checking oversized commands

[Eliot Courtney]

The limit is 16 pages for a single command sent to the GSP. Return an
error if `allocate_command` is called with a too large size.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq.rs                 | 7 ++++++-
 drivers/gpu/nova-core/gsp/fw.rs                   | 4 ++++
 drivers/gpu/nova-core/gsp/fw/r570_144/bindings.rs | 1 +
 3 files changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
index 5cb845bae123..73012da436e5 100644
--- a/drivers/gpu/nova-core/gsp/cmdq.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq.rs
@@ -32,7 +32,8 @@
             GspMsgElement,
             MsgFunction,
             MsgqRxHeader,
-            MsgqTxHeader, //
+            MsgqTxHeader,
+            GSP_MSG_QUEUE_ELEMENT_SIZE_MAX, //
         },
         PteArray,
         GSP_PAGE_SHIFT,
@@ -300,9 +301,13 @@ fn driver_write_area_size(&self) -> usize {
     ///
     /// # Errors
     ///
+    /// - `EMSGSIZE` if the command is larger than [`GSP_MSG_QUEUE_ELEMENT_SIZE_MAX`].
     /// - `ETIMEDOUT` if space does not become available within the timeout.
     /// - `EIO` if the command header is not properly aligned.
     fn allocate_command(&mut self, size: usize, timeout: Delta) -> Result<GspCommand<'_>> {
+        if size_of::<GspMsgElement>() + size > GSP_MSG_QUEUE_ELEMENT_SIZE_MAX {
+            return Err(EMSGSIZE);
+        }
         read_poll_timeout(
             || Ok(self.driver_write_area_size()),
             |available_bytes| *available_bytes >= size_of::<GspMsgElement>() + size,
diff --git a/drivers/gpu/nova-core/gsp/fw.rs b/drivers/gpu/nova-core/gsp/fw.rs
index 4b998485360b..6005362450cb 100644
--- a/drivers/gpu/nova-core/gsp/fw.rs
+++ b/drivers/gpu/nova-core/gsp/fw.rs
@@ -39,6 +39,10 @@
     },
 };
 
+/// Maximum size of a single GSP message queue element in bytes.
+pub(crate) const GSP_MSG_QUEUE_ELEMENT_SIZE_MAX: usize =
+    num::u32_as_usize(bindings::GSP_MSG_QUEUE_ELEMENT_SIZE_MAX);
+
 /// Empty type to group methods related to heap parameters for running the GSP firmware.
 enum GspFwHeapParams {}
 
diff --git a/drivers/gpu/nova-core/gsp/fw/r570_144/bindings.rs b/drivers/gpu/nova-core/gsp/fw/r570_144/bindings.rs
index 6d25fe0bffa9..334e8be5fde8 100644
--- a/drivers/gpu/nova-core/gsp/fw/r570_144/bindings.rs
+++ b/drivers/gpu/nova-core/gsp/fw/r570_144/bindings.rs
@@ -43,6 +43,7 @@ fn fmt(&self, fmt: &mut ::core::fmt::Formatter<'_>) -> ::core::fmt::Result {
 pub const GSP_FW_WPR_META_REVISION: u32 = 1;
 pub const GSP_FW_WPR_META_MAGIC: i64 = -2577556379034558285;
 pub const REGISTRY_TABLE_ENTRY_TYPE_DWORD: u32 = 1;
+pub const GSP_MSG_QUEUE_ELEMENT_SIZE_MAX: u32 = 65536;
 pub type __u8 = ffi::c_uchar;
 pub type __u16 = ffi::c_ushort;
 pub type __u32 = ffi::c_uint;

-- 
2.53.0

[PATCH v5 5/9] gpu: nova-core: gsp: clarify invariant on command queue

[Eliot Courtney]

Clarify why using only the first returned slice from allocate_command
for the message headers is okay.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq.rs | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
index 73012da436e5..c91da368206d 100644
--- a/drivers/gpu/nova-core/gsp/cmdq.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq.rs
@@ -534,7 +534,9 @@ pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
             .gsp_mem
             .allocate_command(command_size, Self::ALLOCATE_TIMEOUT)?;
 
-        // Extract area for the command itself.
+        // Extract area for the command itself. The GSP message header and the command header
+        // together are guaranteed to fit entirely into a single page, so it's ok to only look
+        // at `dst.contents.0` here.
         let (cmd, payload_1) = M::Command::from_bytes_mut_prefix(dst.contents.0).ok_or(EIO)?;
 
         // Fill the header and command in-place.

-- 
2.53.0

[PATCH v5 6/9] gpu: nova-core: gsp: unconditionally call variable payload handling

[Eliot Courtney]

Unconditionally call the variable length payload code, which is a no-op
if there is no such payload but could defensively catch some coding
errors by e.g. checking that the allocated size is completely filled.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq.rs | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
index c91da368206d..48cf28b41f39 100644
--- a/drivers/gpu/nova-core/gsp/cmdq.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq.rs
@@ -548,16 +548,14 @@ pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
             command.init().__init(core::ptr::from_mut(cmd))?;
         }
 
-        // Fill the variable-length payload.
-        if command_size > size_of::<M::Command>() {
-            let mut sbuffer =
-                SBufferIter::new_writer([&mut payload_1[..], &mut dst.contents.1[..]]);
-            command.init_variable_payload(&mut sbuffer)?;
+        // Fill the variable-length payload, which may be empty.
+        let mut sbuffer = SBufferIter::new_writer([&mut payload_1[..], &mut dst.contents.1[..]]);
+        command.init_variable_payload(&mut sbuffer)?;
 
-            if !sbuffer.is_empty() {
-                return Err(EIO);
-            }
+        if !sbuffer.is_empty() {
+            return Err(EIO);
         }
+        drop(sbuffer);
 
         // Compute checksum now that the whole message is ready.
         dst.header

-- 
2.53.0

[PATCH v5 7/9] gpu: nova-core: gsp: add `size_in_bytes` helper to `CommandToGsp`

[Eliot Courtney]

Add a default method to `CommandToGsp` which computes the size of a
command.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq.rs | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
index 48cf28b41f39..3424be4e15f8 100644
--- a/drivers/gpu/nova-core/gsp/cmdq.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq.rs
@@ -94,6 +94,12 @@ fn init_variable_payload(
     ) -> Result {
         Ok(())
     }
+
+    /// Total size of the command (including its variable-length payload) without the
+    /// [`GspMsgElement`] header.
+    fn size_in_bytes(&self) -> usize {
+        size_of::<Self::Command>() + self.variable_payload_len()
+    }
 }
 
 /// Trait representing messages received from the GSP.
@@ -529,10 +535,10 @@ pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
         // This allows all error types, including `Infallible`, to be used for `M::InitError`.
         Error: From<M::InitError>,
     {
-        let command_size = size_of::<M::Command>() + command.variable_payload_len();
+        let size_in_bytes = command.size_in_bytes();
         let dst = self
             .gsp_mem
-            .allocate_command(command_size, Self::ALLOCATE_TIMEOUT)?;
+            .allocate_command(size_in_bytes, Self::ALLOCATE_TIMEOUT)?;
 
         // Extract area for the command itself. The GSP message header and the command header
         // together are guaranteed to fit entirely into a single page, so it's ok to only look
@@ -540,7 +546,7 @@ pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
         let (cmd, payload_1) = M::Command::from_bytes_mut_prefix(dst.contents.0).ok_or(EIO)?;
 
         // Fill the header and command in-place.
-        let msg_element = GspMsgElement::init(self.seq, command_size, M::FUNCTION);
+        let msg_element = GspMsgElement::init(self.seq, size_in_bytes, M::FUNCTION);
         // SAFETY: `msg_header` and `cmd` are valid references, and not touched if the initializer
         // fails.
         unsafe {

-- 
2.53.0

[PATCH v5 8/9] gpu: nova-core: gsp: support large RPCs via continuation record

[Eliot Courtney]

Splits large RPCs if necessary and sends the remaining parts using
continuation records. RPCs that do not need continuation records
continue to write directly into the command buffer. Ones that do write
into a staging buffer first, so there is one copy.

Continuation record for receive is not necessary to support at the
moment because those replies do not need to be read and are currently
drained by retrying `receive_msg` on `ERANGE`.

Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq.rs              |  42 ++++++-
 drivers/gpu/nova-core/gsp/cmdq/continuation.rs | 163 +++++++++++++++++++++++++
 drivers/gpu/nova-core/gsp/fw.rs                |   4 +
 3 files changed, 207 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
index 3424be4e15f8..492e9489e808 100644
--- a/drivers/gpu/nova-core/gsp/cmdq.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq.rs
@@ -1,5 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 
+mod continuation;
+
 use core::{
     mem,
     sync::atomic::{
@@ -25,6 +27,11 @@
     },
 };
 
+use continuation::{
+    ContinuationRecord,
+    SplitState, //
+};
+
 use crate::{
     driver::Bar0,
     gsp::{
@@ -520,7 +527,7 @@ fn notify_gsp(bar: &Bar0) {
             .write(bar);
     }
 
-    /// Sends `command` to the GSP.
+    /// Sends `command` to the GSP, without splitting it.
     ///
     /// # Errors
     ///
@@ -529,7 +536,7 @@ fn notify_gsp(bar: &Bar0) {
     ///   written to by its [`CommandToGsp::init_variable_payload`] method.
     ///
     /// Error codes returned by the command initializers are propagated as-is.
-    pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
+    fn send_single_command<M>(&mut self, bar: &Bar0, command: M) -> Result
     where
         M: CommandToGsp,
         // This allows all error types, including `Infallible`, to be used for `M::InitError`.
@@ -588,6 +595,37 @@ pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
         Ok(())
     }
 
+    /// Sends `command` to the GSP.
+    ///
+    /// The command may be split into multiple messages if it is large.
+    ///
+    /// # Errors
+    ///
+    /// - `ETIMEDOUT` if space does not become available within the timeout.
+    /// - `EIO` if the variable payload requested by the command has not been entirely
+    ///   written to by its [`CommandToGsp::init_variable_payload`] method.
+    ///
+    /// Error codes returned by the command initializers are propagated as-is.
+    pub(crate) fn send_command<M>(&mut self, bar: &Bar0, command: M) -> Result
+    where
+        M: CommandToGsp,
+        Error: From<M::InitError>,
+    {
+        match SplitState::new(command)? {
+            SplitState::Single(command) => self.send_single_command(bar, command),
+            SplitState::Split(command, mut continuations) => {
+                self.send_single_command(bar, command)?;
+
+                while let Some(continuation) = continuations.next() {
+                    // Turbofish needed because the compiler cannot infer M here.
+                    self.send_single_command::<ContinuationRecord<'_>>(bar, continuation)?;
+                }
+
+                Ok(())
+            }
+        }
+    }
+
     /// Wait for a message to become available on the message queue.
     ///
     /// This works purely at the transport layer and does not interpret or validate the message
diff --git a/drivers/gpu/nova-core/gsp/cmdq/continuation.rs b/drivers/gpu/nova-core/gsp/cmdq/continuation.rs
new file mode 100644
index 000000000000..3b83cdcbb39e
--- /dev/null
+++ b/drivers/gpu/nova-core/gsp/cmdq/continuation.rs
@@ -0,0 +1,163 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! Support for splitting large GSP commands across continuation records.
+
+use core::convert::Infallible;
+
+use kernel::prelude::*;
+
+use super::CommandToGsp;
+
+use crate::{
+    gsp::fw::{
+        GspMsgElement,
+        MsgFunction,
+        GSP_MSG_QUEUE_ELEMENT_SIZE_MAX, //
+    },
+    sbuffer::SBufferIter,
+};
+
+/// Maximum command size that fits in a single queue element.
+const MAX_CMD_SIZE: usize = GSP_MSG_QUEUE_ELEMENT_SIZE_MAX - size_of::<GspMsgElement>();
+
+/// Acts as an iterator over the continuation records for a split command.
+pub(super) struct ContinuationRecords {
+    payload: KVVec<u8>,
+    offset: usize,
+}
+
+impl ContinuationRecords {
+    /// Creates a new iterator over continuation records for the given payload.
+    fn new(payload: KVVec<u8>) -> Self {
+        Self { payload, offset: 0 }
+    }
+
+    /// Returns the next continuation record, or [`None`] if there are no more.
+    pub(super) fn next(&mut self) -> Option<ContinuationRecord<'_>> {
+        let remaining = self.payload.len() - self.offset;
+
+        if remaining > 0 {
+            let chunk_size = remaining.min(MAX_CMD_SIZE);
+            let record =
+                ContinuationRecord::new(&self.payload[self.offset..(self.offset + chunk_size)]);
+            self.offset += chunk_size;
+            Some(record)
+        } else {
+            None
+        }
+    }
+}
+
+/// The [`ContinuationRecord`] command.
+pub(super) struct ContinuationRecord<'a> {
+    data: &'a [u8],
+}
+
+impl<'a> ContinuationRecord<'a> {
+    /// Creates a new [`ContinuationRecord`] command with the given data.
+    fn new(data: &'a [u8]) -> Self {
+        Self { data }
+    }
+}
+
+impl<'a> CommandToGsp for ContinuationRecord<'a> {
+    const FUNCTION: MsgFunction = MsgFunction::ContinuationRecord;
+    type Command = ();
+    type InitError = Infallible;
+
+    fn init(&self) -> impl Init<Self::Command, Self::InitError> {
+        <()>::init_zeroed()
+    }
+
+    fn variable_payload_len(&self) -> usize {
+        self.data.len()
+    }
+
+    fn init_variable_payload(
+        &self,
+        dst: &mut SBufferIter<core::array::IntoIter<&mut [u8], 2>>,
+    ) -> Result {
+        dst.write_all(self.data)
+    }
+}
+
+/// Whether a command needs to be split across continuation records or not.
+pub(super) enum SplitState<C: CommandToGsp> {
+    /// A command that fits in a single queue element.
+    Single(C),
+    /// A command split across continuation records.
+    Split(SplitCommand<C>, ContinuationRecords),
+}
+
+impl<C: CommandToGsp> SplitState<C> {
+    /// Maximum variable payload size that fits in the first command alongside the command header.
+    const MAX_FIRST_PAYLOAD: usize = MAX_CMD_SIZE - size_of::<C::Command>();
+
+    /// Creates a new [`SplitState`] for the given command.
+    ///
+    /// If the command is too large, it will be split into a main command and some number of
+    /// continuation records.
+    pub(super) fn new(command: C) -> Result<Self> {
+        let payload_len = command.variable_payload_len();
+
+        if command.size_in_bytes() > MAX_CMD_SIZE {
+            let mut command_payload =
+                KVVec::<u8>::from_elem(0u8, payload_len.min(Self::MAX_FIRST_PAYLOAD), GFP_KERNEL)?;
+            let mut continuation_payload =
+                KVVec::<u8>::from_elem(0u8, payload_len - command_payload.len(), GFP_KERNEL)?;
+            let mut sbuffer = SBufferIter::new_writer([
+                command_payload.as_mut_slice(),
+                continuation_payload.as_mut_slice(),
+            ]);
+
+            command.init_variable_payload(&mut sbuffer)?;
+            if !sbuffer.is_empty() {
+                return Err(EIO);
+            }
+            drop(sbuffer);
+
+            Ok(Self::Split(
+                SplitCommand::new(command, command_payload),
+                ContinuationRecords::new(continuation_payload),
+            ))
+        } else {
+            Ok(Self::Single(command))
+        }
+    }
+}
+
+/// A command that has been truncated to maximum accepted length of the command queue.
+///
+/// The remainder of its payload is expected to be sent using [`ContinuationRecords`].
+pub(super) struct SplitCommand<C: CommandToGsp> {
+    command: C,
+    payload: KVVec<u8>,
+}
+
+impl<C: CommandToGsp> SplitCommand<C> {
+    /// Creates a new [`SplitCommand`] wrapping `command` with the given truncated payload.
+    fn new(command: C, payload: KVVec<u8>) -> Self {
+        Self { command, payload }
+    }
+}
+
+impl<C: CommandToGsp> CommandToGsp for SplitCommand<C> {
+    const FUNCTION: MsgFunction = C::FUNCTION;
+    type Command = C::Command;
+    type InitError = C::InitError;
+
+    fn init(&self) -> impl Init<Self::Command, Self::InitError> {
+        self.command.init()
+    }
+
+    fn variable_payload_len(&self) -> usize {
+        self.payload.len()
+    }
+
+    fn init_variable_payload(
+        &self,
+        dst: &mut SBufferIter<core::array::IntoIter<&mut [u8], 2>>,
+    ) -> Result {
+        dst.write_all(&self.payload)
+    }
+}
diff --git a/drivers/gpu/nova-core/gsp/fw.rs b/drivers/gpu/nova-core/gsp/fw.rs
index 6005362450cb..25fca1f6db2c 100644
--- a/drivers/gpu/nova-core/gsp/fw.rs
+++ b/drivers/gpu/nova-core/gsp/fw.rs
@@ -202,6 +202,7 @@ pub(crate) enum MsgFunction {
     AllocObject = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_OBJECT,
     AllocRoot = bindings::NV_VGPU_MSG_FUNCTION_ALLOC_ROOT,
     BindCtxDma = bindings::NV_VGPU_MSG_FUNCTION_BIND_CTX_DMA,
+    ContinuationRecord = bindings::NV_VGPU_MSG_FUNCTION_CONTINUATION_RECORD,
     Free = bindings::NV_VGPU_MSG_FUNCTION_FREE,
     GetGspStaticInfo = bindings::NV_VGPU_MSG_FUNCTION_GET_GSP_STATIC_INFO,
     GetStaticInfo = bindings::NV_VGPU_MSG_FUNCTION_GET_STATIC_INFO,
@@ -239,6 +240,9 @@ fn try_from(value: u32) -> Result<MsgFunction> {
             bindings::NV_VGPU_MSG_FUNCTION_ALLOC_OBJECT => Ok(MsgFunction::AllocObject),
             bindings::NV_VGPU_MSG_FUNCTION_ALLOC_ROOT => Ok(MsgFunction::AllocRoot),
             bindings::NV_VGPU_MSG_FUNCTION_BIND_CTX_DMA => Ok(MsgFunction::BindCtxDma),
+            bindings::NV_VGPU_MSG_FUNCTION_CONTINUATION_RECORD => {
+                Ok(MsgFunction::ContinuationRecord)
+            }
             bindings::NV_VGPU_MSG_FUNCTION_FREE => Ok(MsgFunction::Free),
             bindings::NV_VGPU_MSG_FUNCTION_GET_GSP_STATIC_INFO => Ok(MsgFunction::GetGspStaticInfo),
             bindings::NV_VGPU_MSG_FUNCTION_GET_STATIC_INFO => Ok(MsgFunction::GetStaticInfo),

-- 
2.53.0

[PATCH v5 9/9] gpu: nova-core: gsp: add tests for continuation records

[Eliot Courtney]

Add tests for continuation record splitting. They cover boundary
conditions at the split points to make sure the right number of
continuation records are made. They also check that the data
concatenated is correct.

Tested-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
---
 drivers/gpu/nova-core/gsp/cmdq/continuation.rs | 138 +++++++++++++++++++++++++
 1 file changed, 138 insertions(+)

diff --git a/drivers/gpu/nova-core/gsp/cmdq/continuation.rs b/drivers/gpu/nova-core/gsp/cmdq/continuation.rs
index 3b83cdcbb39e..637942917237 100644
--- a/drivers/gpu/nova-core/gsp/cmdq/continuation.rs
+++ b/drivers/gpu/nova-core/gsp/cmdq/continuation.rs
@@ -161,3 +161,141 @@ fn init_variable_payload(
         dst.write_all(&self.payload)
     }
 }
+
+#[kunit_tests(nova_core_gsp_continuation)]
+mod tests {
+    use super::*;
+
+    use kernel::transmute::{
+        AsBytes,
+        FromBytes, //
+    };
+
+    /// Non-zero-sized command header for testing.
+    #[repr(C)]
+    #[derive(Clone, Copy, Zeroable)]
+    struct TestHeader([u8; 64]);
+
+    // SAFETY: `TestHeader` is a plain array of bytes for which all bit patterns are valid.
+    unsafe impl FromBytes for TestHeader {}
+
+    // SAFETY: `TestHeader` is a plain array of bytes for which all bit patterns are valid.
+    unsafe impl AsBytes for TestHeader {}
+
+    struct TestPayload {
+        data: KVVec<u8>,
+    }
+
+    impl TestPayload {
+        fn generate_pattern(len: usize) -> Result<KVVec<u8>> {
+            let mut data = KVVec::with_capacity(len, GFP_KERNEL)?;
+            for i in 0..len {
+                // Mix in higher bits so the pattern does not repeat every 256 bytes.
+                data.push((i ^ (i >> 8)) as u8, GFP_KERNEL)?;
+            }
+            Ok(data)
+        }
+
+        fn new(len: usize) -> Result<Self> {
+            Ok(Self {
+                data: Self::generate_pattern(len)?,
+            })
+        }
+    }
+
+    impl CommandToGsp for TestPayload {
+        const FUNCTION: MsgFunction = MsgFunction::Nop;
+        type Command = TestHeader;
+        type InitError = Infallible;
+
+        fn init(&self) -> impl Init<Self::Command, Self::InitError> {
+            TestHeader::init_zeroed()
+        }
+
+        fn variable_payload_len(&self) -> usize {
+            self.data.len()
+        }
+
+        fn init_variable_payload(
+            &self,
+            dst: &mut SBufferIter<core::array::IntoIter<&mut [u8], 2>>,
+        ) -> Result {
+            dst.write_all(self.data.as_slice())
+        }
+    }
+
+    /// Maximum variable payload size that fits in the first command alongside the header.
+    const MAX_FIRST_PAYLOAD: usize = SplitState::<TestPayload>::MAX_FIRST_PAYLOAD;
+
+    fn read_payload(cmd: impl CommandToGsp) -> Result<KVVec<u8>> {
+        let len = cmd.variable_payload_len();
+        let mut buf = KVVec::from_elem(0u8, len, GFP_KERNEL)?;
+        let mut sbuf = SBufferIter::new_writer([buf.as_mut_slice(), &mut []]);
+        cmd.init_variable_payload(&mut sbuf)?;
+        drop(sbuf);
+        Ok(buf)
+    }
+
+    struct SplitTest {
+        payload_size: usize,
+        num_continuations: usize,
+    }
+
+    fn check_split(t: SplitTest) -> Result {
+        let payload = TestPayload::new(t.payload_size)?;
+        let mut num_continuations = 0;
+
+        let buf = match SplitState::new(payload)? {
+            SplitState::Single(cmd) => read_payload(cmd)?,
+            SplitState::Split(cmd, mut continuations) => {
+                let mut buf = read_payload(cmd)?;
+                assert!(size_of::<TestHeader>() + buf.len() <= MAX_CMD_SIZE);
+
+                while let Some(cont) = continuations.next() {
+                    let payload = read_payload(cont)?;
+                    assert!(payload.len() <= MAX_CMD_SIZE);
+                    buf.extend_from_slice(&payload, GFP_KERNEL)?;
+                    num_continuations += 1;
+                }
+
+                buf
+            }
+        };
+
+        assert_eq!(num_continuations, t.num_continuations);
+        assert_eq!(
+            buf.as_slice(),
+            TestPayload::generate_pattern(t.payload_size)?.as_slice()
+        );
+        Ok(())
+    }
+
+    #[test]
+    fn split_command() -> Result {
+        check_split(SplitTest {
+            payload_size: 0,
+            num_continuations: 0,
+        })?;
+        check_split(SplitTest {
+            payload_size: MAX_FIRST_PAYLOAD,
+            num_continuations: 0,
+        })?;
+        check_split(SplitTest {
+            payload_size: MAX_FIRST_PAYLOAD + 1,
+            num_continuations: 1,
+        })?;
+        check_split(SplitTest {
+            payload_size: MAX_FIRST_PAYLOAD + MAX_CMD_SIZE,
+            num_continuations: 1,
+        })?;
+        check_split(SplitTest {
+            payload_size: MAX_FIRST_PAYLOAD + MAX_CMD_SIZE + 1,
+            num_continuations: 2,
+        })?;
+        check_split(SplitTest {
+            payload_size: MAX_FIRST_PAYLOAD + MAX_CMD_SIZE * 3 + MAX_CMD_SIZE / 2,
+            num_continuations: 4,
+        })?;
+        Ok(())
+    }
+}

-- 
2.53.0

Re: [PATCH v5 7/9] gpu: nova-core: gsp: add `size_in_bytes` helper to `CommandToGsp`

[Alexandre Courbot]

On Wed Mar 4, 2026 at 10:42 AM JST, Eliot Courtney wrote:
> Add a default method to `CommandToGsp` which computes the size of a
> command.
>
> Tested-by: Zhi Wang <zhiw@nvidia.com>
> Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
> ---
>  drivers/gpu/nova-core/gsp/cmdq.rs | 12 +++++++++---
>  1 file changed, 9 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
> index 48cf28b41f39..3424be4e15f8 100644
> --- a/drivers/gpu/nova-core/gsp/cmdq.rs
> +++ b/drivers/gpu/nova-core/gsp/cmdq.rs
> @@ -94,6 +94,12 @@ fn init_variable_payload(
>      ) -> Result {
>          Ok(())
>      }
> +
> +    /// Total size of the command (including its variable-length payload) without the
> +    /// [`GspMsgElement`] header.
> +    fn size_in_bytes(&self) -> usize {

We usually don't specify the unit in the method name (the doccomment is
a better place for that). I can fix this when applying.

Re: [PATCH v5 7/9] gpu: nova-core: gsp: add `size_in_bytes` helper to `CommandToGsp`

[Eliot Courtney]

On Wed Mar 4, 2026 at 11:53 AM JST, Alexandre Courbot wrote:
> On Wed Mar 4, 2026 at 10:42 AM JST, Eliot Courtney wrote:
>> Add a default method to `CommandToGsp` which computes the size of a
>> command.
>>
>> Tested-by: Zhi Wang <zhiw@nvidia.com>
>> Signed-off-by: Eliot Courtney <ecourtney@nvidia.com>
>> ---
>>  drivers/gpu/nova-core/gsp/cmdq.rs | 12 +++++++++---
>>  1 file changed, 9 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/gpu/nova-core/gsp/cmdq.rs b/drivers/gpu/nova-core/gsp/cmdq.rs
>> index 48cf28b41f39..3424be4e15f8 100644
>> --- a/drivers/gpu/nova-core/gsp/cmdq.rs
>> +++ b/drivers/gpu/nova-core/gsp/cmdq.rs
>> @@ -94,6 +94,12 @@ fn init_variable_payload(
>>      ) -> Result {
>>          Ok(())
>>      }
>> +
>> +    /// Total size of the command (including its variable-length payload) without the
>> +    /// [`GspMsgElement`] header.
>> +    fn size_in_bytes(&self) -> usize {
>
> We usually don't specify the unit in the method name (the doccomment is
> a better place for that). I can fix this when applying.

Thanks, I will make a note of this so I don't make this mistake again. I
wanted to clarify this since size for a command could mean a few things.

Re: [PATCH v5 2/9] gpu: nova-core: gsp: add mechanism to wait for space on command queue

[Danilo Krummrich]

On Wed Mar 4, 2026 at 2:42 AM CET, Eliot Courtney wrote:
> +    fn allocate_command(&mut self, size: usize, timeout: Delta) -> Result<GspCommand<'_>> {
> +        read_poll_timeout(
> +            || Ok(self.driver_write_area_size()),
> +            |available_bytes| *available_bytes >= size_of::<GspMsgElement>() + size,
> +            Delta::ZERO,

Isn't this either creating unneccessary thrashing of the memory controller or
unnecessary contention at the cache-coherency level?

I think we should probably add at least a small delay of something around 1us.

> +            timeout,
> +        )?;

Claude review: gpu: nova-core: gsp: add continuation record support

[Claude Code Review Bot]

Overall Series Review

Subject: gpu: nova-core: gsp: add continuation record support
Author: Eliot Courtney <ecourtney@nvidia.com>
Patches: 13
Reviewed: 2026-03-05T13:55:43.016540

---

This is a well-structured 9-patch series for the nova-core (Rust NVIDIA GPU driver) that adds continuation record support to the GSP command queue. This allows large RPCs (>16 pages / 64KB) to be split across multiple command queue entries. The series follows a clean incremental approach: preparatory cleanups, infrastructure additions, and then the core feature with tests.

The code quality is generally high, with good documentation, proper error handling, and thorough test coverage. The design with `SplitState` / `SplitCommand` / `ContinuationRecord` is clean and maintains backward compatibility for small commands. I have a few observations but no blocking issues.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: sort `MsgFunction` variants alphabetically

[Claude Code Review Bot]

Patch Review

Straightforward cleanup sorting enum variants and their `TryFrom` match arms alphabetically. The patch also adds `// Common function codes` and `// Event codes` section comments to the `try_from` implementation, which is a nice consistency improvement matching the enum definition.

No issues.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: add mechanism to wait for space on command queue

[Claude Code Review Bot]

Patch Review

This replaces the immediate `EAGAIN` return with a `read_poll_timeout` loop, which is the right approach for supporting continuation records that can fill the queue.

The `driver_write_area_size` method is a good addition that cleanly computes the available space without needing to materialize the slices. The comment explaining the minimum value is helpful.

One observation: the `read_poll_timeout` uses `Delta::ZERO` for the poll interval:

```rust
read_poll_timeout(
    || Ok(self.driver_write_area_size()),
    |available_bytes| *available_bytes >= size_of::<GspMsgElement>() + size,
    Delta::ZERO,
    timeout,
)?;
```

A zero poll interval means busy-waiting. Other callers in the same tree (e.g., `wait_for_msg` at cmdq.rs:572) use `Delta::from_millis(1)`. This could be intentional since command queue space should free up very quickly after GSP processes a command, but it might be worth adding a brief comment explaining the choice or using a small interval to avoid CPU spinning under load.

The timeout of 1 second (`ALLOCATE_TIMEOUT`) matches nouveau's approach, which is good.

---
Generated by Claude Code Patch Reviewer

Claude review: rust: add EMSGSIZE error code

[Claude Code Review Bot]

Patch Review

Trivial and correct. Already has Acked-by and Reviewed-by tags.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: add checking oversized commands

[Claude Code Review Bot]

Patch Review

Adds the `EMSGSIZE` check before attempting allocation. The binding for `GSP_MSG_QUEUE_ELEMENT_SIZE_MAX` (65536 = 16 * 4096) is consistent with the 16-page limit mentioned in the cover letter.

The check:
```rust
if size_of::<GspMsgElement>() + size > GSP_MSG_QUEUE_ELEMENT_SIZE_MAX {
    return Err(EMSGSIZE);
}
```

This is correct - after patch 8 lands, this check becomes effectively dead code for the `allocate_command` path because `SplitState::new` will split commands before they reach `allocate_command` (renamed to `send_single_command`). However, keeping it as a defensive check in `allocate_command` is reasonable as it documents the invariant and catches programming errors.

Minor: the trailing `//` comment on the import line (line 1131 in the mbox) is a `rustfmt` artifact for forcing multi-line formatting. It's consistent with the codebase style.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: clarify invariant on command queue

[Claude Code Review Bot]

Patch Review

Adds a comment explaining why only `dst.contents.0` is used for extracting the command header. This is a documentation-only change.

The comment:
```rust
// Extract area for the command itself. The GSP message header and the command header
// together are guaranteed to fit entirely into a single page, so it's ok to only look
// at `dst.contents.0` here.
```

This relies on the fact that `GspMsgElement` + any `Command` struct together fit in `GSP_PAGE_SIZE` (4096 bytes). This is a reasonable assumption for the fixed-size command headers currently in use, but worth noting this would break if a `Command` struct ever exceeded ~4KB. The invariant is implicitly enforced by the fact that the queue element starts page-aligned and `GspMsgElement::from_bytes_mut_prefix` splits at the element boundary.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: unconditionally call variable payload handling

[Claude Code Review Bot]

Patch Review

Removes the `if command_size > size_of::<M::Command>()` guard around the variable payload handling. This is a good defensive change - calling `init_variable_payload` on zero-length payloads is a no-op by default, and the `sbuffer.is_empty()` check will catch any mismatch between `variable_payload_len()` and actual bytes written.

The explicit `drop(sbuffer)` is needed because `sbuffer` borrows `payload_1` and `dst.contents.1` mutably, and the subsequent checksum computation needs to read them. Good.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: add `size_in_bytes` helper to `CommandToGsp`

[Claude Code Review Bot]

Patch Review

Adds a default method computing total command size. Clean refactor. The method is later used by `SplitState::new` in patch 8, making it a necessary abstraction.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: support large RPCs via continuation record

[Claude Code Review Bot]

Patch Review

This is the main patch. The design is solid:

1. `SplitState::new()` checks if splitting is needed and stages the payload into kernel-allocated buffers
2. `SplitCommand` wraps the original command with truncated payload for the first message
3. `ContinuationRecords` iterates over remaining payload chunks

A few observations:

**`ContinuationRecords` doesn't implement `Iterator`**: The `next()` method signature matches `Iterator` but it doesn't implement the trait. This is likely intentional since `Iterator` would require `&mut self` lifetime complications with the returned `ContinuationRecord<'_>` borrowing from `self`. The usage in `send_command` with `while let Some(...)` is fine. Consider a brief comment explaining this design choice.

**Missing `Tested-by` tag**: This is the only non-trivial patch without a `Tested-by: Zhi Wang` tag (patches 1-7 and 9 all have it). This may be intentional since the patch changed substantively in v4/v5.

**The `SplitState::MAX_FIRST_PAYLOAD` associated constant**:
```rust
const MAX_FIRST_PAYLOAD: usize = MAX_CMD_SIZE - size_of::<C::Command>();
```
This is defined as an associated constant on the generic `SplitState<C>`, which is elegant since it depends on `C::Command`. It's also exposed to tests in patch 9 via `SplitState::<TestPayload>::MAX_FIRST_PAYLOAD`, which is a nice pattern.

**Allocation strategy**: The staging buffer uses `KVVec::from_elem(0u8, ...)` which zero-initializes. The data is immediately overwritten by `init_variable_payload`, so `KVVec::with_capacity` + `resize` would save the zero-fill, but this is a minor optimization concern for a path that's only taken for large (>64KB) RPCs.

**`ContinuationRecord` uses `Command = ()`**: This means the GspMsgElement header is written but the command body is entirely variable-length payload. This matches the continuation record protocol where only the raw data matters.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: add tests for continuation records

[Claude Code Review Bot]

Patch Review

Good test coverage with boundary conditions:
- 0 bytes payload (no split)
- Exactly `MAX_FIRST_PAYLOAD` (no split, boundary)
- `MAX_FIRST_PAYLOAD + 1` (first split)
- `MAX_FIRST_PAYLOAD + MAX_CMD_SIZE` (single continuation, boundary)
- `MAX_FIRST_PAYLOAD + MAX_CMD_SIZE + 1` (two continuations)
- `MAX_FIRST_PAYLOAD + 3.5 * MAX_CMD_SIZE` (multiple continuations)

The test pattern generation with `(i ^ (i >> 8)) as u8` ensures non-repeating patterns across 256-byte boundaries, which is a good approach for catching off-by-one copy errors.

The `read_payload` helper correctly verifies payload content by reading back through the `CommandToGsp` interface.

One minor style note: the `// SAFETY` comments on `TestHeader`'s `FromBytes`/`AsBytes` impls are straightforward since it's `[u8; 64]`. No issues.

Overall this is a well-crafted series ready for merging, with the only substantive note being the zero poll interval choice in patch 2 which may deserve a comment or small delay.

---
Generated by Claude Code Patch Reviewer

Claude review: gpu: nova-core: gsp: clarify invariant on command queue

[Claude Code Review Bot]

Patch Review

Adds a comment explaining why only `dst.contents.0` is used for the command header:

```rust
// Extract area for the command itself. The GSP message header and the command header
// together are guaranteed to fit entirely into a single page, so it's ok to only look
// at `dst.contents.0` here.
```

Good documentation of a non-obvious invariant. No issues.

---
Generated by Claude Code Patch Reviewer