Let user C-c when waiting for DWARF index finalization

In PR gdb/29854, Simon pointed out that it would be good to be able to
use C-c when the DWARF cooked index is waiting for finalization.  The
idea here is to be able to interrupt a command like "break" -- not to
stop the finalization process itself, which runs in a worker thread.

This patch implements this idea, by changing the index wait functions
to, by default, allow a quit.  Polling is done, because there doesn't
seem to be a better way to interrupt a wait on a std::future.

For v2, I realized that the thread compatibility code in thread-pool.h
also needed an update.

Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29854

2 files changed, 21 insertions, 6 deletions

diff --git a/gdb/dwarf2/cooked-index.c b/gdb/dwarf2/cooked-index.c
index f253a42..3d23a65 100644
--- a/gdb/dwarf2/cooked-index.c
+++ b/gdb/dwarf2/cooked-index.c
@@ -28,6 +28,7 @@
 #include <algorithm>
 #include "safe-ctype.h"
 #include "gdbsupport/selftest.h"
+#include <chrono>
 
 /* See cooked-index.h.  */
 
@@ -404,6 +405,21 @@ cooked_index_shard::find (const std::string &name, bool completing) const
   return range (lower, upper);
 }
 
+/* See cooked-index.h.  */
+
+void
+cooked_index_shard::wait (bool allow_quit) const
+{
+  if (allow_quit)
+    {
+      std::chrono::milliseconds duration { 15 };
+      while (m_future.wait_for (duration) == gdb::future_status::timeout)
+	QUIT;
+    }
+  else
+    m_future.wait ();
+}
+
 cooked_index::cooked_index (vec_type &&vec)
   : m_vector (std::move (vec))
 {
diff --git a/gdb/dwarf2/cooked-index.h b/gdb/dwarf2/cooked-index.h
index 7a8216a..dc26956 100644
--- a/gdb/dwarf2/cooked-index.h
+++ b/gdb/dwarf2/cooked-index.h
@@ -259,10 +259,7 @@ public:
   void finalize ();
 
   /* Wait for this index's finalization to be complete.  */
-  void wait () const
-  {
-    m_future.wait ();
-  }
+  void wait (bool allow_quit = true) const;
 
   friend class cooked_index;
 
@@ -373,8 +370,10 @@ public:
        end up writing to freed memory.  Waiting for finalization to
        complete avoids this problem; and the cost seems ignorable
        because creating and immediately destroying the debug info is a
-       relatively rare thing to do.  */
-    wait ();
+       relatively rare thing to do.  Do not allow quitting from this
+       wait.  */
+    for (auto &item : m_vector)
+      item->wait (false);
   }
 
   /* A range over a vector of subranges.  */