mpl: seed standard cells when all macros are fixed

[oharboe]

When every macro is fixed, HierRTLMP::run() skipped the whole pipeline,
including generateTemporaryStdCellsPlacement() and the soft blockages
around macros. A full run seeds every standard cell at the center of its
cluster before global placement reads the database, so a macro placement
re-injected through place_macro/MACRO_PLACEMENT_TCL diverged from the run
that generated it (~2% achieved period measured on a macro-array design)
even with bit-identical macro geometry.

With this PR a design with standard cells goes through clustering,
shaping and the soft-macro annealing even when there are no unfixed
macros: fixed macros act as fixed obstacles (as they already did in mixed
fixed/unfixed designs), no macro is moved, standard cells are seeded at
their cluster centers and the soft blockages are created — matching what
a generating run leaves behind. Macro-only designs with no standard cells
skip as before.

Test: all_fixed_macros1 locks every macro and checks the standard cells
come out placed, no macro moved, and the blockages exist. The
no_unfixed_macros golden gains the design-data report block now printed
before the skip; its DEF is unchanged.

🤖 Generated with Claude Code

[gemini-code-assist]

Code Review

This pull request updates the macro placer to ensure that designs with standard cells but no unfixed macros still go through the clustering process. This allows the temporary standard-cell placement and soft blockages to be generated consistently, which is useful when re-injecting placements. A new integration test all_fixed_macros1 has been added to verify this behavior. There are no review comments, so I have no feedback to provide.