matthew/sonic-buildimage
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[Mellanox] Advance hw-mgmt to v.7.0020.4104 (#13372)

Browse Source 
- Why I did it
Advance hw-mgmt service to V.7.0020.4100
Add missing thermal sensors that are supported by hw-mgmt package
Delay system health service before hw-mgmt has started on Mellanox platform in order to avoid reading some sensors before ready.
Depends on sonic-net/sonic-linux-kernel#305

- How I did it
1. Update hw mgmt version
2. Add missing sensors
3. Delay service 

- How to verify it
Regression test.

Signed-off-by: Stephen Sun <stephens@nvidia.com>
This commit is contained in:
Stephen Sun 2023-02-12 17:23:47 +08:00 committed by mssonicbld
parent 4f3b649f8e
commit b0416a5c2c
11 changed files with 79 additions and 47 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
device/mellanox/x86_64-mlnx_msn2010-r0/platform.json
View File

@ -67,6 +67,9 @@
},
{
"name": "CPU Core 3 Temp"
},
{
"name": "SODIMM 1 Temp"
}
],
"sfps": [

3
device/mellanox/x86_64-mlnx_msn2100-r0/platform.json
View File

@ -67,6 +67,9 @@
},
{
"name": "CPU Core 3 Temp"
},
{
"name": "SODIMM 1 Temp"
}
],
"sfps": [

3
device/mellanox/x86_64-mlnx_msn2410-r0/platform.json
View File

@ -114,6 +114,9 @@
},
{
"name": "CPU Pack Temp"
},
{
"name": "SODIMM 1 Temp"
}
],
"sfps": [

3
device/mellanox/x86_64-mlnx_msn2700-r0/platform.json
View File

@ -114,6 +114,9 @@
},
{
"name": "CPU Pack Temp"
},
{
"name": "SODIMM 1 Temp"
}
],
"sfps": [

3
device/mellanox/x86_64-nvidia_sn2201-r0/platform.json
View File

@ -101,6 +101,9 @@
},
{
"name": "ASIC"
},
{
"name": "SODIMM 1 Temp"
}
],
"sfps": [{

2
platform/mellanox/hw-management.mk
View File

@ -16,7 +16,7 @@
#
# Mellanox HW Management
MLNX_HW_MANAGEMENT_VERSION = 7.0020.3006
MLNX_HW_MANAGEMENT_VERSION = 7.0020.4104
export MLNX_HW_MANAGEMENT_VERSION

10
platform/mellanox/hw-management/0001-Make-SONiC-determine-reboot-cause-service-start-afte.patch
View File

@ -1,8 +1,8 @@
From 1a1011b6da491d35001df5a7204d4eecb2769767 Mon Sep 17 00:00:00 2001
From 489764eb124e03087eb408dec27d769fa4f98459 Mon Sep 17 00:00:00 2001
From: keboliu <kebol@mellanox.com>
Date: Fri, 15 Jan 2021 14:41:16 +0800
Subject: [PATCH] Make SONiC determine-reboot-cause service start after hw-mgmt
service
Subject: [PATCH 1/4] Make SONiC determine-reboot-cause service start after
hw-mgmt service
Signed-off-by: Kebo Liu <kebol@nvidia.com>
---
@ -10,7 +10,7 @@ Signed-off-by: Kebo Liu <kebol@nvidia.com>
1 file changed, 1 insertion(+)
diff --git a/debian/hw-management.hw-management.service b/debian/hw-management.hw-management.service
index 39a2a54..2104b87 100755
index 8bdcaef..1c25ffb 100755
--- a/debian/hw-management.hw-management.service
+++ b/debian/hw-management.hw-management.service
@@ -1,6 +1,7 @@
@ -22,5 +22,5 @@ index 39a2a54..2104b87 100755
[Service]
Type=oneshot
--
1.9.1
2.20.1

52
platform/mellanox/hw-management/0002-Disable-hw-mgmt-on-SimX-platforms.patch
View File

@ -1,59 +1,47 @@
From 79dadd5b0d2f5e860b525c12d4d3843607b03a9f Mon Sep 17 00:00:00 2001
From 422b64397f2f33b394d037820f0ceb4c09e3a725 Mon Sep 17 00:00:00 2001
From: Alexander Allen <arallen@nvidia.com>
Date: Fri, 21 Jan 2022 16:47:19 +0000
Subject: [PATCH] Disable hw-mgmt on SimX platforms
Subject: [PATCH 2/4] Disable hw-mgmt on SimX platforms
---
usr/usr/bin/hw-management-ready.sh | 31 ++++++++++++++++--------------
usr/usr/bin/hw-management-ready.sh | 11 +++++++----
usr/usr/bin/hw-management.sh | 9 +++++++++
2 files changed, 26 insertions(+), 14 deletions(-)
2 files changed, 16 insertions(+), 4 deletions(-)
diff --git a/usr/usr/bin/hw-management-ready.sh b/usr/usr/bin/hw-management-ready.sh
index 5a9698c..364f906 100755
index 88672a8..7558c68 100755
--- a/usr/usr/bin/hw-management-ready.sh
+++ b/usr/usr/bin/hw-management-ready.sh
@@ -51,19 +51,22 @@ if [ -d /var/run/hw-management ]; then
@@ -51,17 +51,20 @@ if [ -d /var/run/hw-management ]; then
rm -fr /var/run/hw-management
fi
-case $board_type in
-VMOD0014)
- while [ ! -d /sys/devices/pci0000:00/0000:00:1f.0/NVSN2201:00/mlxreg-hotplug/hwmon ]
- do
- sleep 1
- done
- ;;
-*)
- while [ ! -d /sys/devices/platform/mlxplat/mlxreg-hotplug/hwmon ]
- do
- sleep 1
- done
- ;;
-esac
+if [ -z "$(lspci -vvv | grep SimX)" ]; then
+ case $board_type in
+ VMOD0014)
+ while [ ! -d /sys/devices/pci0000:00/0000:00:1f.0/NVSN2201:00/mlxreg-hotplug/hwmon ]
+ do
+ sleep 1
+ done
+ ;;
if [ ! -d /sys/devices/pci0000:00/0000:00:1f.0/NVSN2201:00/mlxreg-hotplug/hwmon ]; then
timeout 180 bash -c 'until [ -d /sys/devices/pci0000:00/0000:00:1f.0/NVSN2201:00/mlxreg-hotplug/hwmon ]; do sleep 0.2; done'
fi
;;
-*)
+ *)
+ while [ ! -d /sys/devices/platform/mlxplat/mlxreg-hotplug/hwmon ]
+ do
+ sleep 1
+ done
+ ;;
if [ ! -d /sys/devices/platform/mlxplat/mlxreg-hotplug/hwmon ]; then
timeout 180 bash -c 'until [ -d /sys/devices/platform/mlxplat/mlxreg-hotplug/hwmon ]; do sleep 0.2; done'
fi
;;
-esac
+ esac
+fi
+
echo "Start Chassis HW management service."
logger -t hw-management -p daemon.notice "Start Chassis HW management service."
diff --git a/usr/usr/bin/hw-management.sh b/usr/usr/bin/hw-management.sh
index ebfabb0..c0c038e 100755
index 1ee05b5..50d922b 100755
--- a/usr/usr/bin/hw-management.sh
+++ b/usr/usr/bin/hw-management.sh
@@ -1495,6 +1495,13 @@ do_chip_down()
@@ -2310,6 +2310,13 @@ do_chip_down()
/usr/bin/hw-management-thermal-events.sh change hotplug_asic down %S %p
}
@ -67,7 +55,7 @@ index ebfabb0..c0c038e 100755
__usage="
Usage: $(basename "$0") [Options]
@@ -1520,6 +1527,8 @@ Options:
@@ -2335,6 +2342,8 @@ Options:
force-reload Performs hw-management 'stop' and the 'start.
"
@ -77,5 +65,5 @@ index ebfabb0..c0c038e 100755
start)
if [ -d /var/run/hw-management ]; then
--
2.17.1
2.20.1

8
platform/mellanox/hw-management/0003-Remove-unused-non-upstream-kernel-modules-from-load.patch
View File

@ -1,14 +1,14 @@
From 14b06a12802fc0e15116a64f419d002d0d21d695 Mon Sep 17 00:00:00 2001
From 439639e939f896f9aee42a4dbd5216feb728220c Mon Sep 17 00:00:00 2001
From: Alexander Allen <arallen@nvidia.com>
Date: Thu, 17 Feb 2022 04:19:50 +0000
Subject: [PATCH] Remove unused non-upstream kernel modules from load
Subject: [PATCH 3/4] Remove unused non-upstream kernel modules from load
---
usr/etc/modules-load.d/05-hw-management-modules.conf | 2 --
1 file changed, 2 deletions(-)
diff --git a/usr/etc/modules-load.d/05-hw-management-modules.conf b/usr/etc/modules-load.d/05-hw-management-modules.conf
index 39f621e..c0980bc 100644
index cfcfaa4..dd3b5ca 100644
--- a/usr/etc/modules-load.d/05-hw-management-modules.conf
+++ b/usr/etc/modules-load.d/05-hw-management-modules.conf
@@ -15,8 +15,6 @@ xdpe12284
@ -21,5 +21,5 @@ index 39f621e..c0980bc 100644
gpio-pca953x
pmbus
--
2.17.1
2.20.1

29
platform/mellanox/hw-management/0004-Make-system-health-service-starts-after-hw-managemen.patch Normal file
View File

@ -0,0 +1,29 @@
From 038bce6bf808ec9d082e96fec4184e060b3a85a9 Mon Sep 17 00:00:00 2001
From: Stephen Sun <stephens@nvidia.com>
Date: Mon, 28 Nov 2022 03:55:14 +0000
Subject: [PATCH 4/4] Make system-health service starts after hw-management to
avoid failures
On SN2410, it can fail to read the file led_status_capability if it starts from ONIE
Signed-off-by: Stephen Sun <stephens@nvidia.com>
---
debian/hw-management.hw-management.service | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/debian/hw-management.hw-management.service b/debian/hw-management.hw-management.service
index 1c25ffb..0fbd877 100755
--- a/debian/hw-management.hw-management.service
+++ b/debian/hw-management.hw-management.service
@@ -1,7 +1,7 @@
[Unit]
Description=Chassis HW management service of Mellanox systems
Documentation=man:hw-management.service(8)
-Before=determine-reboot-cause.service
+Before=determine-reboot-cause.service system-health.service
[Service]
Type=oneshot
--
2.20.1

2
platform/mellanox/hw-management/hw-mgmt

@ -1 +1 @@
Subproject commit 137109ed15be147a99b59ceb9dfa1799688dfb71
Subproject commit 9c3f6b2d6ac4be2eda936522f369ff8920deb4e2