268321 - Small cleanup of gfx\src\os2

Status: RESOLVED FIXED

(Core Graveyard :: GFX: OS/2, defect)

Description

[Peter Weilbacher]

When building gfx\src\os2 I get lots of warnings about unused variables,
variables that might be used uninitialized, and missing casts. Nothing is
anywhere critical, but it was difficult to find the real error messages among
all those warnings. ;-)

Comment 1

[Peter Weilbacher]

Attachment: minor cleanup, diff

Comment 2

[Peter Weilbacher]

Comment on attachment 165078 [details] [diff] [review]
minor cleanup, diff

Mike, can you especially have a look if I interpreted the missing casts in
nsImageOS2.cpp correctly? I tested the patch and did not notice any problems
afterwards, but you never know...

Comment 3

[Dainis Jonitis]

Isn't it better to change the PRINTDLG member variables and function parameters
to be consistent, instead of using casts all around?

Comment 4

[Peter Weilbacher]

Attachment: 2nd try, make every printer number ULONG

Hmm, I guess you think about numPrinter and related (mQueueCount etc.)? In this
patch I have converted everything that deals with the number of printers to
ULONG, as this is what SplEnumQueue requires. (Seems that I have to update the
patch to bug 199763 as well which originally led me to do this.)

Comment 5

[Peter Weilbacher]

I only notice now how stupid it is to assign bugs to myself, even if I create
the patches. I cannot check them in myself anyway and then mkaply is not sent
mail to be reminded of reviews and checkin requirements...

Comment 6

[Mike Kaply [:mkaply]]

Comment on attachment 166046 [details] [diff] [review]
2nd try, make every printer number ULONG

r=mkaply, sr=blizzard (platform specific), a=mkaply for 1.8

Comment 7

[Dainis Jonitis]

>      if (!str) {
> -      for (int i = count - 1; i >= 0; i--) 
> +      for (ULONG i = count - 1; i >= 0; i--) 
>          nsMemory::Free(array[i]);

Hey, with ULONG i will never be less than 0! Hang due to infinite loop.
I guess it is better to make this loop to go in opposite direction.
      for (ULONG i = 0 ; i < count ; i++)

Comment 8

[Peter Weilbacher]

Attachment: corrected loop, final patch

Yes, absolutely correct. Stupid me just blindly changed all relevant ints to
ULONGs in the for loops instead of looking closely at what they are doing.

Comment 9

[Peter Weilbacher]

Comment on attachment 166579 [details] [diff] [review]
corrected loop, final patch

Carrying over r/sr/a. Mike, can you check this in please? Or someone else?

Comment 10

[Mike Kaply [:mkaply]]

Fix checked in.

This is trunk only.