Adding platform support for Juniper QFX5200

[ciju-juniper]

This is a 1RU switch with 32 QSFP28 (40G/100G) ports on Broadcom Tomahawk I chipset. CPU used in QFX5200-32C-S is Intel Ivy Bridge. The switch has Redundant and hot-swappable Power Supply (1+1) and also has Redundant and hot swappable fans (5).

Signed-off-by: Ciju Rajan K [email protected]
Signed-off-by: Ashish Bhensdadia [email protected]

- What I did

• Added the device specific files, bcm configuration, portmapping file for Broadcom TH1
• Developed device drivers for platform specific fpga
• Developed the platform scripts

- How I did it
There is a issue #4347 which is being debugged. This patch set has been verified after applying the workarounds/fixes provide on the latest of master branch.

- How to verify it

1. show platform summary
2. show interface status
3. show interface transceiver eeprom
4. sfputil show presence
5. sensors
6. show platform psustatus

The detailed UT was captured in the file qfx5200_unittest_logs.txt
qfx5200_unittest_logs.txt

- Description for the changelog
Adding platform support for Juniper QFX5200

- A picture of a cute animal (not mandatory but encouraged)

[lgtm-com]

This pull request introduces 105 alerts when merging 82ba140216cdfe91521a048f70f2a057b0f9a6a8 into 60b1649 - view on LGTM.com

new alerts:

• 42 for Unused import
• 30 for Unused local variable
• 11 for Use of 'global' at module level
• 8 for Except block handles 'BaseException'
• 6 for Variable defined multiple times
• 2 for Unreachable code
• 2 for 'import *' may pollute namespace
• 2 for __init__ method returns a value
• 1 for Unnecessary pass
• 1 for Result of integer division may be truncated

[ciju-juniper]

@lguohan @jleveque Could you review and merge this platform?

[lguohan]

can you clean up the alerts raised by lgtm?

[ciju-juniper]

@lguohan Alerts are being cleaned up now. Shall upload the files soon.

[lgtm-com]

This pull request introduces 12 alerts when merging 190336a66ae5ae37931fcbca4cca36a2b66f943a into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 3 for Variable defined multiple times
• 2 for 'import *' may pollute namespace
• 1 for Result of integer division may be truncated

[lgtm-com]

This pull request introduces 11 alerts when merging 5636091273fed7da5db1eee2bde0d73121e3e92a into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[lgtm-com]

This pull request introduces 11 alerts when merging 17eaf522c6adfb5995d088b1b260b15ee3c5375d into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[ciju-juniper]

@lguohan We have addressed all the LGTM alerts. Summary is not updated automatically as we edited in the pull request directly. Please consider merging the patchset.

[lgtm-com]

This pull request introduces 11 alerts when merging db02711e45021cf033cafab5fbb3d894cbbaaec8 into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[lgtm-com]

This pull request introduces 11 alerts when merging 242a6c71b92667fff678a06a0adb0395a5acd022 into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[lgtm-com]

This pull request introduces 11 alerts when merging 5220b1d6b9b7c129446466928327ddc6efb85252 into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[lgtm-com]

This pull request introduces 11 alerts when merging 946483a12a30d944b0d343e0dd398a8c77b08d66 into 2a59551 - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[ciju-juniper]

@lguohan Now that #4379 has been committed, could you consider merging this platform?

[ciju-juniper]

@lguohan Rebased the patches to latest sonic-buildimage master.

[lgtm-com]

This pull request introduces 11 alerts when merging 4fa848d1457ba02a21145c8554c73860be060d6b into de377eb - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[lguohan]

it looks like there are still some alerts from lgtm.

[ciju-juniper]

@lguohan A few alerts for nested exception handling. We had marked it as false positives ad these are valid scenarios.

2 Alerts are for "import *". This is needed right now.
2 Alerts are of variable defined. LGTM doesn't understand "os.command" so it's just flagging it as false positive.
Integer division alert is expected to have truncation. We are handling it already.

[ciju-juniper]

@lguohan I've resolved your comments. Also provided the reason why we need adt7470 driver customization for QFX5200.

[lgtm-com]

This pull request introduces 11 alerts when merging 10ef2f3ec2bf988a3ee503c98e2b07cd70efe4ad into 7405f8c - view on LGTM.com

new alerts:

• 6 for Except block handles 'BaseException'
• 2 for 'import *' may pollute namespace
• 2 for Variable defined multiple times
• 1 for Result of integer division may be truncated

[ciju-juniper]

@lguohan We re-organized the code and fixed all the lgtm python alerts. Requesting you to merge this platform.

[ciju-juniper]

@lguohan Could you merge this pull request?

[ciju-juniper]

retest vsimage

[ciju-juniper]

@lguohan We have tested on the latest of master and everything is working as expected. Please do merge the patch set.

[jleveque]

Retest vsimage please

[ciju-juniper]

@jleveque I see that you were able to trigger the vsimage test. I tried it earlier by adding 'retest vsimage' comment, but it didn't trigger. Does it mean only maintainers can trigger it?

I see that last test in vsimage test suite is always failing with most of the pull requests. Are there any issues with the test suite? At least this pull request doesn't touch any of the VS platform files.

[jleveque]

@ciju-juniper: You can also trigger it. The plugin requires you add the word "please" at the end (it forces you to be polite :) ).

I'm not aware of issues with the vsimage test suite, but there may be. I'll see if anyone is aware of this.

[ciju-juniper]

@jleveque Thanks.. I will be more polite from now on ;-)

[ciju-juniper]

@jleveque Would you be able to merge this pull request? There aren't any changes which affects vsimage.

[jleveque]

I will wait for @lguohan to approve and make the decision.

[ciju-juniper]

@lguohan I've refreshed the pull request after removing the adt7470 private driver.

[jleveque]

@ciju-juniper: Please try to use force-pushes only when necessary. Having a history of commits in GitHub make reviewing iterations easier.

[ciju-juniper]

@jleveque I will keep that in mind. I was trying to get a single clean commit in the pull request as it's a new platform addition.

[jleveque]

@ciju-juniper: You don't need to worry about getting a single, clean commit. When we merge pull requests, we use a "squash merge" which will squash all commits in the PR into one single, clean commit.

[ciju-juniper]

@jleveque Thanks for the info.

[ciju-juniper]

@lguohan @jleveque Looks like all the builds/tests passed this time. Would you be able to merge now? Also, I've addressed your comment regarding adt7470 driver.

[ciju-juniper]

@lguohan @jleveque I see that all the checks have passed now including the review. Would this pull request be merged automatically?