mpl: reportLocations prints "invalid format specifier" instead of macro locations

Describe the bug

SimulatedAnnealingCore<T>::reportLocations() formats integer database units with a floating-point
specifier, so every row of the location table fails to format. The table header and rule are
printed, then the body is replaced by a logger error.

This is debug-output only — placement results are unaffected.

Expected Behavior

The location table lists each macro's position and size, in microns, matching the Outline: line
printed immediately above it by HierRTLMP::printPlacementResult.

Environment

Observed at 945a9f48dc6e5cc91d865daa92c45a1094cb682c.

To Reproduce

Any src/mpl/test design, with the placement channel at level 2:

set_debug_level MPL hierarchical_macro_placement 2
source "centralization1.tcl"

Relevant log output

 Cluster Id |                                Location
-----------------------------------------------------

[*** LOG ERROR #0001 ***] [logger] invalid format specifier

Across all 34 designs in src/mpl/test, every one produces this error and no location row is
emitted in any of them.

Cause

In both loops of reportLocations (src/mpl/src/SimulatedAnnealingCore.cpp):

logger_->report("{:>11d} | ({:^8.2f} {:^8.2f}) ({:^8.2f} {:^8.2f})",
                display_id,
                macro.getX(),        // int
                macro.getY(),        // int
                macro.getWidth(),    // int
                macro.getHeight());  // int

getX(), getY(), getWidth() and getHeight() return int, while {:^8.2f} requires a
floating-point argument, so fmt rejects the call.

The intended form appears in HierRTLMP::printPlacementResult, which uses the same specifiers and
converts first — and prints correctly in the same output:

logger_->report("Outline: ({:^8.2f} {:^8.2f}) ({:^8.2f} {:^8.2f})",
                block_->dbuToMicrons(outline.xMin()),
                ...

Outline: (  0.00     0.00  ) ( 302.10   302.40 )     <- same specifiers, converted, prints fine

Suggested fix

Wrap the four values in block_->dbuToMicrons(...) in both loops, matching
printPlacementResult.

A second point in the same function, unverified

The fixed-terminal loop appears to skip one entry:

const int number_of_moveable_macros = static_cast<int>(pos_seq_.size());
for (int i = 0; i < macros_.size(); ++i) {
  if (i <= number_of_moveable_macros) {
    continue;
  }

The movable macros occupy indices 0 .. pos_seq_.size() - 1, so the first fixed terminal is at
index pos_seq_.size(); with <= that entry is skipped and < looks intended.

I have not been able to observe this, since no row prints at all today — flagging it as a reading
of the source rather than something measured, in case it is worth fixing in the same pass.