[matcher] Leave inlining decision on next/prev to compiler

Forcing this inline speeds up most benchmarks by a few percents,
but at least for the Duployan benchmark, it slows it down by 40%.
This is on Apple silicon.

I had noticed that HarfRust was performing faster on this benchmark,
and after lots of investigation, this force-inline seemed to be the
culprit.

src/hb-ot-layout-gsubgpos.hh

diff --git a/src/hb-ot-layout-gsubgpos.hh b/src/hb-ot-layout-gsubgpos.hh
index 2c47931..0437e44 100644
--- a/src/hb-ot-layout-gsubgpos.hh
+++ b/src/hb-ot-layout-gsubgpos.hh
@@ -577,9 +577,6 @@
     return SKIP;
   }
 
-#ifndef HB_OPTIMIZE_SIZE
-  HB_ALWAYS_INLINE
-#endif
   bool next (unsigned *unsafe_to = nullptr)
   {
     auto *info = c->buffer->info;
@@ -608,9 +605,6 @@
       *unsafe_to = end;
     return false;
   }
-#ifndef HB_OPTIMIZE_SIZE
-  HB_ALWAYS_INLINE
-#endif
   bool prev (unsigned *unsafe_from = nullptr)
   {
     auto *out_info = c->buffer->out_info;