On 29/4/24 14:31, Richard Henderson wrote:
On 4/29/24 05:18, Philippe Mathieu-Daudé wrote:
+
+    t = ldq_he_p(buf) | ldq_he_p(buf + len - 8);

Here we read #0 and #31, ...

+    p = QEMU_ALIGN_PTR_DOWN(buf + 8, 8);
+    e = QEMU_ALIGN_PTR_DOWN(buf + len - 1, 8);
+
+    /* Read 0 to 31 aligned words from the middle. */

... so here is #1 to #30?

Not indexes, but count.  There may be zero words remaining in the middle, etc.

Oh, got it, thanks!


Reply via email to