On Thu, 5 Nov 2026 18:58:06 GMT, Alan Bateman <[email protected]> wrote:
>> Thanks for the explanation. It makes sense.
>>
>> I find it hard to see what cause the error condition. Maybe it's better to
>> change the comment to
>>
>>
>> } else {
>> // Either of the above GetFinalPathNameByHandleW() calls has failed
>> ...
>
> Here's an alternative that avoids the else branch. This should handle the
> potential for the final path length to change between the two calls.
>
> - if (len >= size) {
> - if ((finalPath = (WCHAR*)malloc(len * sizeof(WCHAR))) == NULL)
> + if (len != 0 && len >= size) {
> + DWORD pathBufferSize = len;
> + finalPath = (WCHAR*)malloc((size_t)pathBufferSize *
> sizeof(WCHAR));
> + if (finalPath == NULL) {
> + CloseHandle(h);
> return NULL;
> - len = GetFinalPathNameByHandleW(h, finalPath, len, 0);
> + }
> + len = GetFinalPathNameByHandleW(h, finalPath, pathBufferSize, 0);
> + if (len == 0 || len >= pathBufferSize) {
> + CloseHandle(h);
> + free(finalPath);
> + return NULL;
> + }
Hi Alan, thanks for the suggestions!
I added them.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/33752#discussion_r4490355076
