* [DELL][Z9100,S6100,S6000] Platform 2.0 SFP Changes
Added support in sfp.py file which will be generic. Send the eeprom path and sfp_control path from chassis.py
Present: Servers are listed in the same order as in redis-db
Fix: Save the sort o/p, hence use sorted list to write into pam.d's conf.
As well convert priority to integer for use by sort.
* Enable debug image build for kvm image.
* Fix a bug in cleaning up requests referring the netsmp_session being closed to avoid a crash, during netsnmp_session close/abort due to timeout.
RCA:
The netsnmp_agent_session requests use netsmp_subtree objects that matches the associated variable name.
The netsnmp_subtrees created through a netsnmp_session are tied to that session.
When a subagent connection is closed/dropped due to timeout, all associated netsnmp_subtree objects are fred.
Hence during a netsnmp_session close, all the delegated netsnmp_agent_sessions are scanned for requets that could be using netsnmp_subtree objects associated with this netsnmp_sesssion or its subsession. For each of the found request, they are explicitly marked to fail and a call is made to complete them.
But due to the bug in scanning, it leaves behind requests and hence later when the requests get processed, they refer the *now* freed netsnmp_subtree. As often these requests gets completed pretty soon, they escape crashing. But if it so happens that the freed memory happened to complete a memory unit, hence returned to kernel or it got reallocated & changed enough to crash, the snmpd crashes.
* Revert the changes
* Revert
* [service dependent] describe non-warm-reboot dependency outside systemctl
When dependency was described with systemctl, it will kick in all the time,
including under warm reboot/restart scenarios. This is not what we always
want. For components that are capable of warm reboot/start, they need to
describe dependency in service files.
Signed-off-by: Ying Xie <ying.xie@microsoft.com>
* [service] teamd service should not require swss service
Adding require swss will cause teamd to be killed by systemctl when swss
stops. This is not what we want in warm reboot.
Signed-off-by: Ying Xie <ying.xie@microsoft.com>
* refactoring code
* rename functions to match other functions in the file
* [sonic-cfg] [Marvell] get_system_mac() to fetch valid mac address from Eeprom else
get mac address from eth0
* [platform] Marvell mac address fetch for different platform based on board
Signed-off-by: Antony Rheneus <arheneus@marvell.com>
* fix sfpd initialize issue
* fix review comments
* rephrase the output log
* fix retry counter
* change the retry time to 10, means set max waiting time 1024s
* fix mlnx-sfpd init flow with new solution
* [mlnx-sfpd] address comments
1. wait for 5 seconds * 30 times, 150 seconds totally. use constant wait time for each retry.
2. use try/except structure so that error can be handled in a graceful way
* [mlnx-sfpd] wait 5 seconds after SDK_DAEMON_READY_FILE exists to make sure SDK is fully up.
* [mlnx-sfpd]simplify initialization by using deinitialize on initializing failure
This switch has 27 fiber ports, 4x25G, 20x10G , and 3x100G ports ports.
CPU: Intel ® Atom™ Processor C3508,1.6GHz
BMC: None
MAC: Broadcom BCM88470 (Qumran AX).
MISC: Support IEEE1588v2, hot-swappable PSU, hot-swappable fan tray.
But notice here, BCM88470 is not supported for SAI now.
So the syncd container is not running so far.
Signed-off-by: roy_lee <roy_lee@accton.com>
Corrected the ingress and egress lossy and lossless buffer pool and profile values.
Single pool for lossy and lossless traffic for PFC priority 3 and 4.
In HWSKU DellEMC-S5232f-P-100GG/25G/10G. Defaults to lossy profile. No Lossless support.
2 default mmu_init.data files (TD3-DEFAULT-LOSSLESS-P3P4 and "TD3-DEFAULT" created in SAI code. This will have cpu pool configuration and MMU init related configurations.
TD3-DEFAULT.data file is only supports lossy.
TD3-DEFAULT-LOSSLESS-P3P4.data file is supports lossless on P3 and P4.
* Utilize the same sai.profile and config.bcm file
Signed-off-by: Zhenggen Xu <zxu@linkedin.com>
* Fix the index issues for HWSKU: Seastone-DX010-25-50 and Seastone-DX010-50-40
Signed-off-by: Zhenggen Xu <zxu@linkedin.com>
* Update sonic-swss submodule
* src/sonic-swss 63afbd5...5be3963 (11):
> Remove the platform check for remove/add port action at init phase. (#1006)
> [test]: Add Class object in each test script file (#992)
> [vstest]: add --imgname option to specify vs image to test (#996)
> [qosorch]: Add cir and pir parameters for scheduler (#991)
> [vstest]: report error when team kernel module is not available (#994)
> Support per buffer pool watermark polling mode (#973)
> [test]: Update test_vlan to use dvs functions (#986)
> Add synchronous switch to orch agent (#987)
> ARM32 bit fixes, for 64bit printf format specifier (#989)
> [portsorch]: Rename function isPortReady() to allPortsReady() (#990)
> [routeorch]: Fix observer detaching procedure for next hop observers (#985)
Signed-off-by: Zhenggen Xu <zxu@linkedin.com>
Add fan control for as7116-54x
Modify sfputil.py to support sfp insert and remove events for as7116-54x
Modify port_config.nps to set port default admin down
Modify port_config.ini for add speed attr
Code format optimization like remove newline
Modify buffers_defaults_t0 and t1, change ingress_lossless_pool mode from dynamic to static
Add nephos_dac.nps and nephos_opt.nps for support difference dac and fiber module
when device disk is small, do not unzip dockerfs.tar.gz on disk.
keep the tar file on the disk, unzip to tmpfs in the initrd phase.
enabled this for 7050-qx32
Signed-off-by: Guohan Lu <gulv@microsoft.com>
* [Makefile/slave docker] ARM arch doesn't support few packages
iproute2 is missing for ARM sonic-slave docker
Signed-off-by: Antony Rheneus <arheneus@marvell.com>
Sflow Tables (#299)
[Warnings] Fix all printf format specifiers for 32bit arch compilation
(#298)
[schema]: Add APP_ACL_TABLE and APP_ACL_RULE_TABLE (#275)
* Adding --key option to sonic-cfggen script. This will help to display config DB with more granularity.
Signed-off-by: Vasant Patil <vapatil@linkedin.com>
Add platform support for delta ag9032v2a
CPU: Broadwell-DE
ASIC: BMC56870 with <100Gx32 + 10Gx1>
BMC: AST2520:
- What I did
add the new platform of delta ag9032v2a
- How I did it
1.provide the QSFP and SWPLD attributes and create the virtual buses of I2C
2. provide the port configuration
- How to verify it
1. psuutil.py
2. sfputil.py
3. bcmcd
- Known issue
The port LED is not ready. We noticed that BCM chip had the M0 FW initialization issue while bringing up SDK with "BRCM SAI ver: [3.5.2.3], OCP SAI ver: [1.4], SDK ver: [6.5.14]" and here is the information :
root@sonic:/home/admin# bcmcmd "M0 status"
M0 status
0:soc_iproc_data_send_wait: No response for msg 2
M0 FW is NOT Running
M0 FW Version is 0.0
Host FW Version is 1.0
Host and M0 FW Versions do not match!!!
Signed-off-by: hans-tseng <hans.tseng@deltaww.com>
xcvrd is broken because a newly introduced interface get_transceiver_dom_threshold_info_dict in common code calls a unsupported interface _read_eeprom_specific_bytes.
Fix the issue by implement get_transceiver_dom_threshold_info_dict to avoid calling the unsupported interface.
The race condition could happen like this:
When an interface is enslaved into the port channel immediately after
it is created, the order of creating the ifinfo and linking the ifinfo to
the port is not guaranteed.
Please check the patch commit message to get full details.
Signed-off-by: Shu0T1an ChenG <shuche@microsoft.com>
I2c burst read may failed due to misoperation of ir3570a(A DC-to-DC converter IC).
As #2966, there are 3 more models have this symptom, as7326-56x, as7726-32x, and as9716-32d.
Also correct typo of naming on as7816-64x and as7716-32x.
Signed-off-by: roy_lee roy_lee@accton.com
What I did
Disabling i2c function of ir3570a which may failed i2c tranfer to others.
Close channel of mux after data transfered.
How I did it
Identify version of ir3570, if it's ir3570a, disable its alias i2c address.
Enable parameter of driver i2c_mux_pca954x to close channel on after every access.
How to verify it
Write 08 to offset 0xcf of systom eeprom and execute i2c block read.
It will return error.
plug-in several transceivers and run "show interfaces transceiver presence" and "sfputil show presence". You may see the different result. (But doing this requires updated xcvrd).
Signed-off-by: roy_lee <roy_lee@accton.com>
* src/sonic-utilities ee56d54...cb0e745 (11):
> sonic_utilities: Support for DOM Threshold values for EEPROM dump
(#545)
> [portstat] Fix portstat show RX_UTIL over 100% for 100G (#563)
> sonic_installer: fix read-only filesystem support for firmware
update (#565)
> Revert "show acl table command output should show binding column
correctly even with single port (#447)" (#589)
> show acl table command output should show binding column correctly
even with single port (#447)
> [config] Do no stop or restart dependent services (#582)
> sfpshow: prevent 'show int trans eeprom --dom' from crashing (#567)
> [warm-reboot] add docker upgrade --warm option and roll back support
(#559)
> [ecnconfig] Validate input WRED parameters (#579)
> [sonic-utilities] Add fstrim to reboot (#535)
> Fixing the expected neighbor command due to change in output format
under sonic-buildimage/pull/3036 (#584)
