1. Add test cases for dynamic buffer calculation
  - changing speed/cable length test
  - headroom override
  - additional lossless PG
  - exceeding the accumulative headroom size of a port
  - default buffer model
2. Adjust qos test, fetching configuration from:
  - APPL_DB if there is buffer_model defined, this is the case in master
  - CONF_DB otherwise, for 201911
3. Additional checking script to test whether the buffer model is as expected.
   This can be  used by warm-reboot

Signed-off-by: Stephen Sun <stephens@nvidia.com>

use pytest_assert and google style doc strings
verify buffer profile and pool against ASIC_DB
move hardcoded parameters to a predefined json file
add comments to make it easy to understand

Signed-off-by: Stephen Sun <stephens@nvidia.com>

Signed-off-by: Stephen Sun <stephens@nvidia.com>

Signed-off-by: Stephen Sun <stephens@nvidia.com>

Signed-off-by: Stephen Sun <stephens@nvidia.com>

Signed-off-by: Stephen Sun <stephens@nvidia.com>

It's more convenent to leave it as a class member for users

Signed-off-by: Stephen Sun <stephens@nvidia.com>

- Use capital letters for global variable names
- Remove spaces between "=" in argument list

Signed-off-by: Stephen Sun <stephens@nvidia.com>

Signed-off-by: Stephen Sun <stephens@nvidia.com>