[Mono-devel-list] Corlib test cases

Sat Jan 24 09:08:48 EST 2004

On Sat, Jan 24, 2004 at 09:29:50AM +0000, David Sheldon wrote:
> On Thu, Jan 22, 2004 at 11:43:31PM +0000, David Sheldon wrote:
> > I have been trying to assign bug numbers to the tests that I can see
> > failing. 
> 
> Another 10,

20 this time, as a lot of them appear to be duplicated.

> 30) MonoTests.System.IO.FileTest.MoveException9 : Expected: IOException but was ArgumentException
>                           
> in [0x000bf] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/System.IO/File.cs:232) System.IO.File:Move (string,string)
> in [0x0004b] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:386) MonoTests.System.IO.FileTest:MoveException9 ()

Interestingly, I would have said that this test was a case where "the source
and destination are the same.", which is stated in the msdn docs to not throw
an exception, yet it does on MS.NET. I guess they are the same file, but not 
the same string.

http://bugs.ximian.com/show_bug.cgi?id=53290

> 31) MonoTests.System.IO.FileTest.CreationTime : test#01 
> 	expected:<2002>
> 	 but was:<2004>
>                           
> in [0x0006f] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:621) MonoTests.System.IO.FileTest:CreationTime ()

Unix doesn't support CreationTime.

> 32) MonoTests.System.IO.FileTest.FileStreamCloseException : IOException was expected

The same Unix idiom difference as number 29. 

> 33) MonoTests.System.IO.FileTest.SetCreationTimeIOException1 : Expected: IOException but was ArgumentOutOfRangeException

in [0x00026] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/System/DateTime.cs:1058) System.DateTime:ToFileTime ()
in [0x0001e] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/System.IO/File.cs:296) System.IO.File:SetCreationTime (string,System.DateTime)
in [0x00047] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:1071) MonoTests.System.IO.FileTest:SetCreationTimeIOException1 ()

Same as above. Windows locks files down completely.  I am really not convinced
that we should be duplicating this behaviour. Are we trying to duplicate Windows, 
or just a runtime?

> 34) MonoTests.System.IO.FileTest.SetCreationTimeUtcIOException1 : Expected: IOException but was ArgumentOutOfRangeException
[snip]
> in [0x00047] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:1144) MonoTests.System.IO.FileTest:SetCreationTimeUtcIOException1 ()

> 35) MonoTests.System.IO.FileTest.SetLastAccessTimeIOException1 : Expected: IOException but was ArgumentOutOfRangeException
> [snip]                          
> in [0x00047] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:1219) MonoTests.System.IO.FileTest:SetLastAccessTimeIOException1 ()
>                           
> 36) MonoTests.System.IO.FileTest.SetLastAccessTimeUtcIOException1 : Expected: IOException but was ArgumentOutOfRangeException
> [snip]                          
> in [0x00047] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:1292) MonoTests.System.IO.FileTest:SetLastAccessTimeUtcIOException1 ()

> 37) MonoTests.System.IO.FileTest.SetLastWriteTimeIOException1 : Expected: IOException but was ArgumentOutOfRangeException
> [snip]
> in [0x00047] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:1367) MonoTests.System.IO.FileTest:SetLastWriteTimeIOException1 ()
>                           
> 38) MonoTests.System.IO.FileTest.SetLastWriteTimeUtcIOException1 : Expected: IOException but was ArgumentOutOfRangeException
[snip]                           
> in [0x00047] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.IO/FileTest.cs:1440) MonoTests.System.IO.FileTest:SetLastWriteTimeUtcIOException1 ()

More of the same.

> 39) MonoTests.System.MathTest.TestAtan : 1.570796326794897 != 1.570796326794897
>                       
> in [0x000d8] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System/MathTest.cs:158) MonoTests.System.MathTest:TestAtan ()
> 
> 40) MonoTests.System.MathTest.TestSinh : 0.12371341868561378 != 0.12371341868561378
>                       
> in [0x0004d] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System/MathTest.cs:247) MonoTests.System.MathTest:TestSinh ()

These appear to be problems in accuracy. The value of double.Epsilon seems to
be insufficient, as the values are out by about 6E-17, whereas double.Epsilon
is 5E-324. Is Epsilon being used incorrectly? 

> 41) MonoTests.System.Reflection.Emit.AssemblyBuilderTest.TestGetFiles1 : NotSupportedException was expected

Added an override for GetFiles() that throws the exception.

> 42) MonoTests.System.Reflection.Emit.AssemblyBuilderTest.TestSetCustomAttribute : System.IO.FileNotFoundException : Could not find file "foo"
>                               
> in [0x0013e] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/System.IO/FileStream.cs:130) System.IO.FileStream:.ctor (string,System.IO.FileMode,System.IO.FileAccess,System.IO.FileShare,int,bool)
> in [0x00018] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/System.IO/FileStream.cs:68) System.IO.FileStream:.ctor (string,System.IO.FileMode)
> in <0x00065> (wrapper remoting-invoke-with-check) System.IO.FileStream:.ctor (string,System.IO.FileMode)
> in [0x00072] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/System.Reflection.Emit/AssemblyBuilder.cs:609) System.Reflection.Emit.AssemblyBuilder:SetCustomAttribute (System.Reflection.Emit.CustomAttributeBuilder)
> in [0x00081] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.Reflection.Emit/AssemblyBuilderTest.cs:402) MonoTests.System.Reflection.Emit.AssemblyBuilderTest:TestSetCustomAttribute ()

http://bugs.ximian.com/show_bug.cgi?id=53292

43-46) - These were test cases requiring Test/resources/ to be resources/ when
they were run. I was running the tests from make run-test, and they were not
being found. Maybe the code should be more inteligent about finding them, or
just be written to work from where the makefile will run them.

> 47) MonoTests.System.Runtime.Remoting.SoapServicesTest.TestGetXmlType : E7 
> 	String lengths differ.  Expected length=86, but was length=159.
> 	Strings differ at index 86.
> 	
> 	expected:<"...System.Runtime.Remoting/corlib_test">
> 	 but was:<"...System.Runtime.Remoting/corlib_test%2C%20Version%3D0.0.0.0%2C%20Cultur...">
> 	------------------------------------------------^
> 	
>                               
> in [0x0009c] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.Runtime.Remoting/SoapServicesTest.cs:103) MonoTests.System.Runtime.Remoting.SoapServicesTest:TestGetXmlType ()

http://bugs.ximian.com/show_bug.cgi?id=53294

I have suggested a patch to this. Will it break anything major? (on the bug)

> 48) MonoTests.System.Runtime.Remoting.SoapServicesTest.TestGetInteropType : T1 
> 	expected:<MonoTests.System.Runtime.Remoting.SoapTest2>
> 	 but was:<(null)>
>                               
> in [0x000e4] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.Runtime.Remoting/SoapServicesTest.cs:158) MonoTests.System.Runtime.Remoting.SoapServicesTest:TestGetInteropType ()

> 49) MonoTests.System.Runtime.Remoting.SoapServicesTest.TestSoapFields : #1 
> 	expected:<"atribut">
> 	 but was:<(null)>
>                               
> in [0x00028] (at /home/sheldon/hacking/mono/raw_cvs/mcs/class/corlib/Test/System.Runtime.Remoting/SoapServicesTest.cs:181) MonoTests.System.Runtime.Remoting.SoapServicesTest:TestSoapFields ()

I think these were also fixed by: http://bugs.ximian.com/show_bug.cgi?id=53294

-- 
[ Gorman orders the troops to unload all their 
weapons before the first alien encounter. ]
Frost: What the hell are we supposed to use, man? Harsh language?