Modified mapper to handle disconnected lif components connected to sa…

[ysingh7]

…me cproc

Issue Number:

Objective of pull request:

Pull request checklist

Your PR fulfills the following requirements:

• Issue created that explains the change and why it's needed
• Tests are part of the PR (for bug fixes / features)
• Docs reviewed and added / updated if needed (for bug fixes / features)
• PR conforms to Coding Conventions
• PR applys BSD 3-clause or LGPL2.1+ Licenses to all code files
• Lint (flakeheaven lint src/lava tests/) and (bandit -r src/lava/.) pass locally
• Build tests (pytest) passes locally

Pull request type

Please check your PR type:

• Bugfix
• Feature
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no api changes)
• Build related changes
• Documentation changes
• Other (please describe):

What is the current behavior?

What is the new behavior?

Does this introduce a breaking change?

• Yes
• No

Supplemental information

[mgkwill]

Hi Yash,

It seems like you are replacing l_addr.core_id in map_cores() and _set_virtual_address_nc() with self.mapper_core_offset that starts at 0 and is incriminated after each use, and kept track of in self.mapper_core_dict.

Later chip_idx and core_idx are calculated based on this replaced value and used in initialization of NcVirtualAddress:

chip_idx = l_addr.core_id // num_cores
core_idx = l_addr.core_id % num_cores

It's not clear to me if l_addr.core_id is entirely arbitrary and can be replaced with an arbitrary value. Can you explain the reasoning behind this?

[ysingh7]

Hi Yash,

It seems like you are replacing l_addr.core_id in map_cores() and _set_virtual_address_nc() with self.mapper_core_offset that starts at 0 and is incriminated after each use, and kept track of in self.mapper_core_dict.

Later chip_idx and core_idx are calculated based on this replaced value and used in initialization of NcVirtualAddress:

chip_idx = l_addr.core_id // num_cores
core_idx = l_addr.core_id % num_cores

It's not clear to me if l_addr.core_id is entirely arbitrary and can be replaced with an arbitrary value. Can you explain the reasoning behind this?

l_addr.core_id is something generated by compiler for a NeuroProcGroup. Two different disconnected NeuroProcGroups can have same exact l_addr.core. So in that sense its arbitary. Job of the mapper is to provide unique id. Thats why we are reassigning the l_core_id.