Sync from internal Google development.

Many improvements, too many to mention.  One significant
perf regression warrants investigation:

omitfp.parsetoproto2_googlemessage1.upb_jit: 343 -> 252 (-26.53)
plain.parsetoproto2_googlemessage1.upb_jit: 334 -> 251 (-24.85)

25% regression for this benchmark is bad, but since I don't think
there's any fundamental design issue that caused it I'm going to
go ahead with the commit anyway.  Can investigate and fix later.

Other benchmarks were neutral or showed slight improvement.

1 files changed, 3 insertions, 6 deletions

diff --git a/upb/bytestream.c b/upb/bytestream.c
index 812e552..8feb678 100644
--- a/upb/bytestream.c
+++ b/upb/bytestream.c
@@ -32,8 +32,6 @@ upb_byteregion *upb_byteregion_newl(const void *str, size_t len) {
   memcpy(ptr, str, len);
   ptr[len] = '\0';
   upb_stringsrc_reset(src, ptr, len);
-  upb_byteregion_fetch(upb_stringsrc_allbytes(src));
-  assert(len == upb_byteregion_available(upb_stringsrc_allbytes(src), 0));
   return upb_stringsrc_allbytes(src);
 }
 
@@ -93,10 +91,10 @@ static upb_stdio_buf *upb_stdio_findbuf(const upb_stdio *s, uint64_t ofs) {
 
 static upb_stdio_buf *upb_stdio_rotatebufs(upb_stdio *s) {
   upb_stdio_buf **reuse = NULL;  // XXX
-  uint32_t num_reused = 0, num_inuse = 0;
+  int num_reused = 0, num_inuse = 0;
 
   // Could sweep only a subset of bufs if this was a hotspot.
-  for (uint32_t i = 0; i < s->nbuf; i++) {
+  for (int i = 0; i < s->nbuf; i++) {
     upb_stdio_buf *buf = s->bufs[i];
     if (buf->refcount > 0) {
       s->bufs[num_inuse++] = buf;
@@ -243,10 +241,9 @@ upb_bytesink* upb_stdio_bytesink(upb_stdio *stdio) { return &stdio->sink; }
 
 upb_bytesuccess_t upb_stringsrc_fetch(void *_src, uint64_t ofs, size_t *read) {
   upb_stringsrc *src = _src;
-  assert(ofs <= src->len);
+  assert(ofs < src->len);
   if (ofs == src->len) {
     upb_status_seteof(&src->bytesrc.status);
-    *read = 0;
     return UPB_BYTE_EOF;
   }
   *read = src->len - ofs;