thr3ads.net - Nouveau - [PATCH v2 13/12] nova-core: sequencer: Refactor run() to handle unknown messages [Nov 2025]

If this information is useful, please help other people find it:
Share via:

Timur Tabi

2025-Nov-03 19:12 UTC

[PATCH v2 00/12] nova-core: Complete GSP boot and begin RPC communication

On Sun, 2025-11-02 at 18:59 -0500, Joel Fernandes wrote:> Hello!
> These patches a refresh of the series adding support for final stages of
the
> GSP boot process where a sequencer which inteprets firmware instructions
needs
> to run to boot the GSP processor, followed by waiting for an INIT_DONE
message
> from the GSP.
> 
> The patches are based on Alex's github branch which have several
prerequisites:
> Repo: https://github.com/Gnurou/linux.git?Branch: b4/gsp_boot
> 
> I also dropped several patches (mainly from John that have already been
> applied).? Tested on Ampere GA102. We also need the "gpu: nova-core:
Add
> get_gsp_info() command" patch which I dropped since it needs to be
reworked,
> and it is not needed for GSP boot on Ampere (but John mentioned it is
needed
> for Blackwell so we could include it in the Blackwell series or I can try
to
> include it in this series if I'm respinning).
I applied your patches on top of Alex's tree, and when I boot on a GA102 I
get this:

[  376.316679] NovaCore 0000:65:00.0: NVIDIA (Chipset: GA102, Architecture:
Ampere, Revision: a.1)
[  377.188060] NovaCore 0000:65:00.0: GSP RPC: send: seq# 0,
function=Ok(GspSetSystemInfo),
length=0x3f0
[  377.188070] NovaCore 0000:65:00.0: GSP RPC: send: seq# 1,
function=Ok(SetRegistry), length=0xc5
[  378.315960] NovaCore 0000:65:00.0: GSP RPC: receive: seq# 0, function=NOCAT,
length=0x50c
[  378.319875] NovaCore 0000:65:00.0: probe with driver NovaCore failed with
error -34

Are you sure there are no other patches?  The RPC patches can't depend on
INIT_DONE being the first
response.  Getting a NOCAT RPC first is not uncommon.

Joel Fernandes

2025-Nov-03 19:35 UTC

head link

[PATCH v2 00/12] nova-core: Complete GSP boot and begin RPC communication

On 11/3/2025 2:12 PM, Timur Tabi wrote:> On Sun, 2025-11-02 at 18:59 -0500, Joel Fernandes wrote:
>> Hello!
>> These patches a refresh of the series adding support for final stages
of the
>> GSP boot process where a sequencer which inteprets firmware
instructions needs
>> to run to boot the GSP processor, followed by waiting for an INIT_DONE
message
>> from the GSP.
>>
>> The patches are based on Alex's github branch which have several
prerequisites:
>> Repo: https://github.com/Gnurou/linux.git?Branch: b4/gsp_boot
>>
>> I also dropped several patches (mainly from John that have already been
>> applied).? Tested on Ampere GA102. We also need the "gpu:
nova-core: Add
>> get_gsp_info() command" patch which I dropped since it needs to be
reworked,
>> and it is not needed for GSP boot on Ampere (but John mentioned it is
needed
>> for Blackwell so we could include it in the Blackwell series or I can
try to
>> include it in this series if I'm respinning).
> 
> I applied your patches on top of Alex's tree, and when I boot on a
GA102 I get this:
> 
> [  376.316679] NovaCore 0000:65:00.0: NVIDIA (Chipset: GA102, Architecture:
Ampere, Revision: a.1)
> [  377.188060] NovaCore 0000:65:00.0: GSP RPC: send: seq# 0,
function=Ok(GspSetSystemInfo),
> length=0x3f0
> [  377.188070] NovaCore 0000:65:00.0: GSP RPC: send: seq# 1,
function=Ok(SetRegistry), length=0xc5
> [  378.315960] NovaCore 0000:65:00.0: GSP RPC: receive: seq# 0,
function=NOCAT, length=0x50c
> [  378.319875] NovaCore 0000:65:00.0: probe with driver NovaCore failed
with error -34
> 
> Are you sure there are no other patches?  The RPC patches can't depend
on INIT_DONE being the first
> response.  Getting a NOCAT RPC first is not uncommon.
It works on my end. Do you have "the wait for init done" patch (the
12th patch?)
You can also boot my tree which has all the patches, i.e. this series +
Alex's
b4/gsp_boot branch:
https://web.git.kernel.org/pub/scm/linux/kernel/git/jfern/linux.git/log/?h=nova-seq-init-done-submitted-v2

[    4.672205] virtme-ng-init: initialization done
[   25.745799] NovaCore 0000:00:07.0: NVIDIA (Chipset: GA102, Architecture:
Ampere, Revision: a.1)
[   26.364343] NovaCore 0000:00:07.0: GSP RPC: send: seq# 0,
function=Ok(GspSetSystemInfo), length=0x3f0
[   26.364634] NovaCore 0000:00:07.0: GSP RPC: send: seq# 1,
function=Ok(SetRegistry), length=0xc5
[   27.561186] NovaCore 0000:00:07.0: GSP RPC: receive: seq# 0,
function=RUN_CPU_SEQUENCER, length=0x18e8
[   27.635180] NovaCore 0000:00:07.0: GSP RPC: receive: seq# 0, function=NOCAT,
length=0x50c
[   27.635529] NovaCore 0000:00:07.0: GSP RPC: receive: seq# 0,
function=LIBOS_PRINT, length=0x68
[   27.635795] NovaCore 0000:00:07.0: GSP RPC: receive: seq# 0,
function=LIBOS_PRINT, length=0x70
[   27.790175] NovaCore 0000:00:07.0: GSP RPC: receive: seq# 0,
function=INIT_DONE, length=0x50

Joel Fernandes

2025-Nov-04 23:26 UTC

head link

[PATCH v2 13/12] nova-core: sequencer: Refactor run() to handle unknown messages

Refactor GspSequencer::run() to follow the same pattern as gsp_init_done()
by wrapping message reception in a loop that ignores unknown messages
(ERANGE errors).

Suggested-by: Timur Tabi <ttabi at nvidia.com>
Signed-off-by: Joel Fernandes <joelagnelf at nvidia.com>
---
Additional patch to cure probe issue on Timur's GA102 (which happens to
receive
too many NOCAT records).

 drivers/gpu/nova-core/gsp/sequencer.rs | 86 +++++++++++++++-----------
 1 file changed, 49 insertions(+), 37 deletions(-)

diff --git a/drivers/gpu/nova-core/gsp/sequencer.rs
b/drivers/gpu/nova-core/gsp/sequencer.rs
index ecc80f668dc8..b98e5146abd8 100644
--- a/drivers/gpu/nova-core/gsp/sequencer.rs
+++ b/drivers/gpu/nova-core/gsp/sequencer.rs
@@ -35,8 +35,8 @@ impl MessageFromGsp for fw::rpc_run_cpu_sequencer_v17_00 {
 
 const CMD_SIZE: usize = size_of::<fw::GSP_SEQUENCER_BUFFER_CMD>();
 
-struct GspSequencerInfo<'a> {
-    info: &'a fw::rpc_run_cpu_sequencer_v17_00,
+struct GspSequencerInfo {
+    cmd_index: u32,
     cmd_data: KVec<u8>,
 }
 
@@ -125,7 +125,7 @@ pub(crate) fn size_bytes(&self) -> usize {
 }
 
 pub(crate) struct GspSequencer<'a> {
-    seq_info: GspSequencerInfo<'a>,
+    seq_info: GspSequencerInfo,
     bar: &'a Bar0,
     sec2_falcon: &'a Falcon<Sec2>,
     gsp_falcon: &'a Falcon<Gsp>,
@@ -368,7 +368,7 @@ fn into_iter(self) -> Self::IntoIter {
         GspSeqIter {
             cmd_data,
             current_offset: 0,
-            total_cmds: self.seq_info.info.cmdIndex,
+            total_cmds: self.seq_info.cmd_index,
             cmds_processed: 0,
             dev: self.dev,
         }
@@ -387,41 +387,53 @@ pub(crate) struct GspSequencerParams<'a> {
 
 impl<'a> GspSequencer<'a> {
     pub(crate) fn run(cmdq: &mut Cmdq, params:
GspSequencerParams<'a>, timeout: Delta) -> Result {
-        cmdq.receive_msg_from_gsp(timeout, |info, mut sbuf| {
-            let cmd_data = sbuf.flush_into_kvec(GFP_KERNEL)?;
-            let seq_info = GspSequencerInfo { info, cmd_data };
-
-            let sequencer = GspSequencer {
-                seq_info,
-                bar: params.bar,
-                sec2_falcon: params.sec2_falcon,
-                gsp_falcon: params.gsp_falcon,
-                libos_dma_handle: params.libos_dma_handle,
-                gsp_fw: params.gsp_fw,
-                dev: params.dev,
-            };
-
-            dev_dbg!(params.dev, "Running CPU Sequencer commands\n");
-
-            for cmd_result in &sequencer {
-                match cmd_result {
-                    Ok(cmd) => cmd.run(&sequencer)?,
-                    Err(e) => {
-                        dev_err!(
-                            params.dev,
-                            "Error running command at index {}\n",
-                            sequencer.seq_info.info.cmdIndex
-                        );
-                        return Err(e);
-                    }
+        let seq_info = loop {
+            match cmdq.receive_msg_from_gsp(
+                timeout,
+                |info: &fw::rpc_run_cpu_sequencer_v17_00, mut sbuf| {
+                    let cmd_data = sbuf.flush_into_kvec(GFP_KERNEL)?;
+                    Ok(GspSequencerInfo {
+                        cmd_index: info.cmdIndex,
+                        cmd_data,
+                    })
+                },
+            ) {
+                Ok(seq_info) => break seq_info,
+                Err(ERANGE) => continue,
+                Err(e) => return Err(e),
+            }
+        };
+
+        let sequencer = GspSequencer {
+            seq_info,
+            bar: params.bar,
+            sec2_falcon: params.sec2_falcon,
+            gsp_falcon: params.gsp_falcon,
+            libos_dma_handle: params.libos_dma_handle,
+            gsp_fw: params.gsp_fw,
+            dev: params.dev,
+        };
+
+        dev_dbg!(params.dev, "Running CPU Sequencer commands\n");
+
+        for cmd_result in &sequencer {
+            match cmd_result {
+                Ok(cmd) => cmd.run(&sequencer)?,
+                Err(e) => {
+                    dev_err!(
+                        params.dev,
+                        "Error running command at index {}\n",
+                        sequencer.seq_info.cmd_index
+                    );
+                    return Err(e);
                 }
             }
+        }
 
-            dev_dbg!(
-                params.dev,
-                "CPU Sequencer commands completed successfully\n"
-            );
-            Ok(())
-        })
+        dev_dbg!(
+            params.dev,
+            "CPU Sequencer commands completed successfully\n"
+        );
+        Ok(())
     }
 }
-- 
2.34.1

Joel Fernandes

2025-Nov-06 23:11 UTC

head link

[PATCH v3 00/14] nova-core: Complete GSP boot and begin RPC communication

Hello!
These patches a refresh of the series adding support for final stages of the
GSP boot process where a sequencer which inteprets firmware instructions needs
to run to boot the GSP processor, followed by waiting for an INIT_DONE message
from the GSP and finally retrieving GPU information from the GSP.

The patches are based on Alex's github branch which have several
prerequisites:
Repo: https://github.com/Gnurou/linux.git Branch: b4/gsp_boot

The entire tree (these patches + patches from Alex, Alistair, John) can be
found at the git repository:
git: git://git.kernel.org/pub/scm/linux/kernel/git/jfern/linux.git
tag: nova-sequencer-init-done-v3

I also dropped several patches (mainly from John that have already been
applied).  Tested on my Ampere GA102 and I see in dmesg:
NovaCore 0000:00:07.0: GPU name: NVIDIA GeForce RTX 3090 Ti

Changes from v2 to v3:
- Added several tags.
- Fixed commit messages, style errors.
- Added GspStaticInfo patch.
- Fixed bug found by Timur in the sequencer code (related to ignoring messages).
- Removed excessive dev_dbg prints in sequencer code (John Hubbard).

Previous series:
v1: https://lore.kernel.org/all/20250829173254.2068763-1-joelagnelf at
nvidia.com/
v2: https://lore.kernel.org/all/20251102235920.3784592-1-joelagnelf at
nvidia.com/

Alistair Popple (2):
  gpu: nova-core: gsp: Wait for gsp initialization to complete
  gpu: nova-core: gsp: Retrieve GSP static info to gather GPU
    information

Joel Fernandes (12):
  gpu: nova-core: falcon: Move waiting until halted to a helper
  gpu: nova-core: falcon: Move start functionality into separate helper
  gpu: nova-core: falcon: Move mbox functionalities into helper
  gpu: nova-core: falcon: Move dma_reset functionality into helper
  gpu: nova-core: gsp: Add support for checking if GSP reloaded
  gpu: nova-core: Add bindings required by GSP sequencer
  gpu: nova-core: Implement the GSP sequencer
  gpu: nova-core: sequencer: Add register opcodes
  gpu: nova-core: sequencer: Add delay opcode support
  gpu: nova-core: sequencer: Implement basic core operations
  gpu: nova-core: sequencer: Implement core resume operation
  gpu: nova-core: sequencer: Refactor run() to handle unknown messages

 drivers/gpu/nova-core/falcon.rs               | 101 +++--
 drivers/gpu/nova-core/falcon/gsp.rs           |  17 +
 drivers/gpu/nova-core/gsp.rs                  |   1 +
 drivers/gpu/nova-core/gsp/boot.rs             |  35 +-
 drivers/gpu/nova-core/gsp/cmdq.rs             |   1 -
 drivers/gpu/nova-core/gsp/commands.rs         | 102 ++++-
 drivers/gpu/nova-core/gsp/fw.rs               |  47 ++
 .../gpu/nova-core/gsp/fw/r570_144/bindings.rs | 248 +++++++++++
 drivers/gpu/nova-core/gsp/sequencer.rs        | 404 ++++++++++++++++++
 drivers/gpu/nova-core/nova_core.rs            |   1 +
 drivers/gpu/nova-core/regs.rs                 |   6 +
 drivers/gpu/nova-core/sbuffer.rs              |   1 -
 drivers/gpu/nova-core/util.rs                 |  16 +
 13 files changed, 943 insertions(+), 37 deletions(-)
 create mode 100644 drivers/gpu/nova-core/gsp/sequencer.rs
 create mode 100644 drivers/gpu/nova-core/util.rs


base-commit: 8adec6f12a71641402fdc77e7d514ceee019312b
-- 
2.34.1

Nouveau - Nov 2025 - [PATCH v2 13/12] nova-core: sequencer: Refactor run() to handle unknown messages

[PATCH v2 00/12] nova-core: Complete GSP boot and begin RPC communication

[PATCH v2 00/12] nova-core: Complete GSP boot and begin RPC communication

[PATCH v2 13/12] nova-core: sequencer: Refactor run() to handle unknown messages

[PATCH v3 00/14] nova-core: Complete GSP boot and begin RPC communication