bugfix m_fillmissingyears macro

[flohump]

Please fill following information
(Add additional info if you think its important and not covered by this Pull Request (PR)):

Purpose of this PR

• bugfix m_fillmissingyears macro. The macro was running over "t" not "t_all". This error-prone. For instance, if t is 2050 and 2100, but data is only availabe until 2095, the macro would use data from 2050 for 2100. Now, the macro runs over t_all, which makes sure that the data from 2095 is used for 2100.

Performance loss/gain from current default behavior

• No test runs performed

Type of change

• Bug fix (Change which fixes an issue).
• New feature (Change which adds functionality).
• Minor change (Change which does not modify core model code i.e., in /modules).
• Major change (fix or feature that would change current model behavior).

How Has This Been Tested?

• gams compilation OK "gams main.gms action=C"
• local test run OK "gams main.gms"

Additions or Changes to default configuration (default.cfg):

Additions are the introduction of new model components in default config

Changes are deletion or updates to the existing model components in default config

• Realizations:
• Scenario switches:
• Scalars/Constants:
• Model interfaces:
• Others:

Checklist:

• Self-review of my own code.
• Compilation check (model starts without compilation errors).
• Added changes to CHANGELOG.md
• No hard coded numbers and cluster/country/region names.
• The code doesn't contain declared but unused parameters or variables.
• In-code comments added including documentation comments.
• Compiled and checked resulting documentation with goxygen::goxygen() for the new/updated parts.
• Changes to magpie4 R library for post processing of model output (ideally backward compatible).

Special comments/warnings

• Notes

[tscheypidi]

looks good