unix: workaround gcc bug on armv7 (#4564)

Disable optimization on `uv__preadv_or_pwritev`.

Fixes: https://github.com/libuv/libuv/issues/4532
Fixes: https://github.com/libuv/libuv/issues/4550

README.md

@@ -343,10 +343,6 @@ after the process terminates unless the event loop is closed.
 
 Use the `ipcrm` command to manually clear up System V resources.
 
-## Known Issues
-
-- A possible arm-linux-gnueabihf-gcc bug causing, sometimes, incorrect generated code on `armv7` when calling `preadv()`: https://github.com/libuv/libuv/issues/4532.
-
 ## Patches
 
 See the [guidelines for contributing][].

src/unix/fs.c

@@ -461,7 +461,12 @@ static ssize_t uv__pwritev_emul(int fd,
 
 /* The function pointer cache is an uintptr_t because _Atomic void*
  * doesn't work on macos/ios/etc...
+ * Disable optimization on armv7 to work around the bug described in
+ * https://github.com/libuv/libuv/issues/4532
  */
+#if defined(__arm__) && (__ARM_ARCH == 7)
+__attribute__((optimize("O0")))
+#endif
 static ssize_t uv__preadv_or_pwritev(int fd,
                                      const struct iovec* bufs,
                                      size_t nbufs,
@@ -482,10 +487,7 @@ static ssize_t uv__preadv_or_pwritev(int fd,
     atomic_store_explicit(cache, (uintptr_t) p, memory_order_relaxed);
   }
 
-  /* Use memcpy instead of `f = p` to work around a compiler bug,
+  f = p;
-   * see https://github.com/libuv/libuv/issues/4532
-   */
-  memcpy(&f, &p, sizeof(p));
   return f(fd, bufs, nbufs, off);
 }