422776 - reduce narrow windows API calls in imagelib

Status: RESOLVED FIXED

(Core :: General, defect)

Description

[Brad Lassey [:blassey] (use needinfo?)]

Attachment: directy from "other" patch on bug 418703

Comment 1

[Brian Crowder]

Review ping?

Comment 2

[Bernard Alleysson]

you have:

+      PRUnichar specialNativePath[MAX_PATH];
+      ::SHGetPathFromIDListW(idList, specialNativePath);
+      ::GetShortPathNameW(specialNativePath, specialNativePath, sizeof(specialNativePath));

Please see:
http://msdn.microsoft.com/en-us/library/aa364989(VS.85).aspx

I think that the patch should use sizeof(specialNativePath)/sizeof(specialNativePath[0]) to have the count of chars (cch).

Please note that NS_ARRAY_LENGTH could be used here and in other places. It is defined here:
http://mxr.mozilla.org/mozilla-central/source/xpcom/glue/nsMemory.h

Comment 3

[Brian Crowder]

Attachment: v2

Updates some buffer issues (as Bernard noticed) and brings in line with trunk.

Comment 4

[Bernard Alleysson]

(In reply to comment #3)
> Created an attachment (id=334779) [details]
> v2

Looks good.

While you're touching that file, you might consider fixing bug 284156 along with it. You'd just have to do:

-      IMalloc* pMalloc;
-      hr = ::SHGetMalloc(&pMalloc);
-      if (SUCCEEDED(hr)) {
-        pMalloc->Free(idList);
-        pMalloc->Release();
-      }
+      ::CoTaskMemFree(idList);

This is very safe, as SHGetMalloc is deprecated per MSDN.

Comment 5

[Brian Crowder]

http://hg.mozilla.org/mozilla-central/rev/e17e0d80a962

Comment 6

[Alexander L. Slovesnik]

Can patch for this bug cause Bug 454247?

Comment 7

[Brian Crowder]

Certainly possible, would you care to try rolling it back?

Comment 8

[Alexander L. Slovesnik]

(In reply to comment #7)
> Certainly possible, would you care to try rolling it back?
Unfortunately I don't have windows build environment, I'm using linux.