February 2009 - Wine-Devel - list.winehq.org

Re: gdiplus: in GdipDrawString only NULL terminate if we have room in the buffer.

by Nikolay Sivov

Aric Stewart wrote: > --- > dlls/gdiplus/graphics.c | 3 ++- > 1 files changed, 2 insertions(+), 1 deletions(-) > > diff --git a/dlls/gdiplus/graphics.c b/dlls/gdiplus/graphics.c > index 2673042..c276a43 100644 > --- a/dlls/gdiplus/graphics.c > +++ b/dlls/gdiplus/graphics.c > @@ -1901,7 +1901,8 @@ GpStatus WINGDIPAPI GdipDrawString(GpGraphics *graphics, GDIPCONST WCHAR *string > j++; > } > > - stringdup[j] = 0; > + if (j < length) > + stringdup[j] = 0; > length = j; > > while(sum < length){ > Hm, maybe I don't follow something, but I think 'j < length' is always true after that: --- for(i = 0, j = 0; i < length; i++){ if(!isprintW(string[i]) && (string[i] != '\n')) continue; stringdup[j] = string[i]; j++; } stringdup[j] = 0; length = j; --- Or you thought about zero length case? It should be handled earlier, on stringdup allocation or before...

Feb. 12, 2009

RE: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Ge van Geldorp

> From: Paul Vriens [mailto:paul.vriens.wine(a)gmail.com] > > Paul Vriens wrote: > > Juan Lang wrote: > >> This removes the unused variables in the last version. > >> --Juan > >> > >> > >> > --------------------------------------------------------------------- > >> --- > >> > >> > > I think that these failures are also because of "not enough rights" > > (people running winetest as a normal user). So just removing them is > > probably not the best approach? > > Maybe it's not that there are not enough rights. The common > denominator that I see is that all reports/boxes that produce > these failures seems to be coming from one person (Ge). Maybe > worthwhile checking with him before just removing the tests? All the failure cases on my machines are when running as non-Administrator. I've set up a number of boxes deliberately that way, but I'm starting to wonder if that is a good idea? For a whole bunch of tests (MSI comes to mind) the only sensible approach is to skip the test when running as non-Administrator, so it seems that I'm not actually adding new test runs in a different environment as I originally wanted to do. Gé.

Feb. 12, 2009

Re: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Reece Dunn

2009/2/12 Paul Vriens <paul.vriens.wine(a)gmail.com>: > Juan Lang wrote: >>>> I think that these failures are also because of "not enough rights" >>>> (people running winetest as a normal user). So just removing them is >>>> probably not the best approach? >>>> >>> Maybe it's not that there are not enough rights. The common denominator that >>> I see is that all reports/boxes that produce these failures seems to be >>> coming from one person (Ge). Maybe worthwhile checking with him before just >>> removing the tests? >> >> Perhaps. The last error didn't change in any of these tests, so I'm >> discinclined to believe it's a permissions issue. The only change was >> that on some systems the return value is FALSE, whereas on others it's >> TRUE. I don't think a test that says ok(ret == TRUE || ret == FALSE) >> is worth much, so I removed it. >> >> In general, I treat this Windows boxes as oracles: if they behave a >> certain way, then by definition that's allowed behavior. The only >> exception is when the behavior differs and is clearly broken on older >> machines. Since in this case, the failures are on relatively new >> machines as well, the tests appear to be flawed. >> >> Of course, if it's a misconfiguration, I'll reconsider. >> --Juan >> > What we maybe need in the header of the report is more information: > > - The locale > - The screen dimensions > - Do we have admin rights +1 - DPI setting - Installed theme (if any) - system metrics (font name & size, etc) -- this would be useful for those tests - installed versions of system components -- MSI; IE; .NET -- I know that these could probably be implied from DLL versions, but it can be difficult to find out what version of a DLL corresponds to IE6SP1 or whatever - Reece

Feb. 12, 2009

Re: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Paul Vriens

Juan Lang wrote: >> What we maybe need in the header of the report is more information: >> >> - The locale >> - The screen dimensions >> - Do we have admin rights >> >> (just brainstorming here). >> >> That would make it easier to decide if it's a misconfiguration or a real >> acceptable configuration. > > Maybe you're right. To that list I'd add the video driver, as the > DirectX folks appear to have black-balled most video cards. > --Juan > Well all cards are fine as long as it's a NVIDIA :) -- Cheers, Paul.

Feb. 12, 2009

Re: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Juan Lang

> What we maybe need in the header of the report is more information: > > - The locale > - The screen dimensions > - Do we have admin rights > > (just brainstorming here). > > That would make it easier to decide if it's a misconfiguration or a real > acceptable configuration. Maybe you're right. To that list I'd add the video driver, as the DirectX folks appear to have black-balled most video cards. --Juan

Feb. 12, 2009

Re: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Paul Vriens

Juan Lang wrote: >>> I think that these failures are also because of "not enough rights" >>> (people running winetest as a normal user). So just removing them is >>> probably not the best approach? >>> >> Maybe it's not that there are not enough rights. The common denominator that >> I see is that all reports/boxes that produce these failures seems to be >> coming from one person (Ge). Maybe worthwhile checking with him before just >> removing the tests? > > Perhaps. The last error didn't change in any of these tests, so I'm > discinclined to believe it's a permissions issue. The only change was > that on some systems the return value is FALSE, whereas on others it's > TRUE. I don't think a test that says ok(ret == TRUE || ret == FALSE) > is worth much, so I removed it. > > In general, I treat this Windows boxes as oracles: if they behave a > certain way, then by definition that's allowed behavior. The only > exception is when the behavior differs and is clearly broken on older > machines. Since in this case, the failures are on relatively new > machines as well, the tests appear to be flawed. > > Of course, if it's a misconfiguration, I'll reconsider. > --Juan > What we maybe need in the header of the report is more information: - The locale - The screen dimensions - Do we have admin rights (just brainstorming here). That would make it easier to decide if it's a misconfiguration or a real acceptable configuration. -- Cheers, Paul.

Feb. 12, 2009

Re: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Juan Lang

>> I think that these failures are also because of "not enough rights" >> (people running winetest as a normal user). So just removing them is >> probably not the best approach? >> > Maybe it's not that there are not enough rights. The common denominator that > I see is that all reports/boxes that produce these failures seems to be > coming from one person (Ge). Maybe worthwhile checking with him before just > removing the tests? Perhaps. The last error didn't change in any of these tests, so I'm discinclined to believe it's a permissions issue. The only change was that on some systems the return value is FALSE, whereas on others it's TRUE. I don't think a test that says ok(ret == TRUE || ret == FALSE) is worth much, so I removed it. In general, I treat this Windows boxes as oracles: if they behave a certain way, then by definition that's allowed behavior. The only exception is when the behavior differs and is clearly broken on older machines. Since in this case, the failures are on relatively new machines as well, the tests appear to be flawed. Of course, if it's a misconfiguration, I'll reconsider. --Juan

Feb. 12, 2009

Re: wintrust(3/3): Fix a couple tests on a variety of systems

by Paul Vriens

Juan Lang wrote: > Hi Paul, > >> This is of course on boxes where people dont' run winetest as a >> administrator. I'm not sure if ERROR_ACCESS_DENIED is thus actually an >> acceptable failure. In other tests we have used skip() for this. > > Of course that's the reason, but not running as administrator is an > acceptable configuration, is it not? I use skip() where subsequent > tests will fail as a result of an early failure, but that wasn't the > case here. > --Juan > I guess the main problem lies in the fact that we as Wine are now also accepting ERROR_ACCESS_DENIED and we won't be noticed of this fact. A skip (which is most of the time valid) at least shows us something. -- Cheers, Paul.

Feb. 12, 2009

Re: wintrust(3/3): Fix a couple tests on a variety of systems

by Juan Lang

Hi Paul, > This is of course on boxes where people dont' run winetest as a > administrator. I'm not sure if ERROR_ACCESS_DENIED is thus actually an > acceptable failure. In other tests we have used skip() for this. Of course that's the reason, but not running as administrator is an acceptable configuration, is it not? I use skip() where subsequent tests will fail as a result of an early failure, but that wasn't the case here. --Juan

Feb. 12, 2009

Re: wintrust(2/3): Remove a couple tests that fail on a variety of systems (try 2)

by Paul Vriens

Paul Vriens wrote: > Juan Lang wrote: >> This removes the unused variables in the last version. >> --Juan >> >> >> ------------------------------------------------------------------------ >> >> > I think that these failures are also because of "not enough rights" > (people running winetest as a normal user). So just removing them is > probably not the best approach? > Maybe it's not that there are not enough rights. The common denominator that I see is that all reports/boxes that produce these failures seems to be coming from one person (Ge). Maybe worthwhile checking with him before just removing the tests? -- Cheers, Paul.

Feb. 12, 2009