wininet: HTTP_QUERY_RAW_HEADERS_CRLF length shouldn't include nul

Dan Kegel dank at kegel.com
Wed Jul 23 18:48:39 CDT 2008


This time with the patch.

On Wed, Jul 23, 2008 at 4:26 PM, Dan Kegel <dank at kegel.com> wrote:
> QueryHttpInfoW's reported length was off by one for
> HTTP_QUERY_RAW_HEADERS_CRLF.
> Add test to check whether HTTP_QUERY_RAW_HEADERS_CRLF
> should include the terminating NUL in the reported length,
> fix problem it exposed.  The code now looks more like neighboring cases.
> Tested on XP.
>
> I started looking at this because it popped up in valgrind runs
> of Picasa.  It also shows up in my daily scans of the conformance test suite.
>
> Fixes Valgrind warnings like:
>  Use of uninitialised value of size 4
>   at  wcstombs_sbcs (wctomb.c:187)
>   by  wine_cp_wcstombs (wctomb.c:457)
>   by  WideCharToMultiByte (locale.c:1936)
>   by  HttpQueryInfoA (http.c:2560)
>   by  HttpHeaders_test (http.c:1158)
>  Uninitialised value was created by a client request
>   at  mark_block_uninitialized (heap.c:164)
>   by  RtlAllocateHeap (heap.c:1239)
>   by  HttpQueryInfoA (http.c:2546)
>   by  HttpHeaders_test (http.c:1158)
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wininet-http-query-crlf.patch
Type: text/x-patch
Size: 2114 bytes
Desc: not available
Url : http://www.winehq.org/pipermail/wine-patches/attachments/20080723/1ba20c76/attachment.bin 


More information about the wine-patches mailing list