Re: (resend)usp10/test: test ScriptXtoX on an RTL set with differring cChars and cGlyphs
Hi,
While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13681
Your paranoid android.
=== WNT4WSSP6 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W2KPROSP4 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== WXPPROSP3 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W2K3R2SESP2 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== WVISTAADM (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W2K8SE (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W7PRO (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W7PROX64 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
I don't get this.
I do the test locally, they all pass. I submit it directly to testbot and they all pass. I submit the patch and i get these nonsense failures.
"should return piX=0 not 0"? seems like 0 == 0 to me. They are both int values...
Is this just testbot being strange?
-aric
On 8/24/11 4:56 PM, Marvin wrote:
Hi,
While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13681
Your paranoid android.
=== WNT4WSSP6 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W2KPROSP4 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== WXPPROSP3 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W2K3R2SESP2 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== WVISTAADM (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W2K8SE (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W7PRO (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
=== W7PROX64 (32 bit usp10) === usp10.c:1931: Test failed: ScriptCPtoX trailing: iCP=8 should return piX=0 not 0
On Thu, 2011-08-25 at 07:15 -0500, Aric Stewart wrote:
"should return piX=0 not 0"? seems like 0 == 0 to me. They are both int values...
That test doesn't look right:
winetest_ok(piX == offsets[iCP+1], "ScriptCPtoX trailing: iCP=%d should return piX=%d not %d\n", iCP, offsets[iCP+direction], piX);
direction can be either -1 or 1.
Yes, it depends on if it is a RTL or LTR string. That is correct.
-aric
On 8/25/11 7:30 AM, Hans Leidekker wrote:
On Thu, 2011-08-25 at 07:15 -0500, Aric Stewart wrote:
"should return piX=0 not 0"? seems like 0 == 0 to me. They are both int values...
That test doesn't look right:
winetest_ok(piX == offsets[iCP+1], "ScriptCPtoX trailing: iCP=%d should return piX=%d not %d\n", iCP, offsets[iCP+direction], piX);
direction can be either -1 or 1.
On Thu, 2011-08-25 at 07:38 -0500, Aric Stewart wrote:
Yes, it depends on if it is a RTL or LTR string. That is correct.
What I meant is that it does not always match the right hand operand in the test condition.
Ah i see what you are seeing.
I will see if that helps. However it is still strange that the tests are all working for me and when i submit tests but not when it is submitted as a patch.
-aric
On 8/25/11 7:47 AM, Hans Leidekker wrote:
On Thu, 2011-08-25 at 07:38 -0500, Aric Stewart wrote:
Yes, it depends on if it is a RTL or LTR string. That is correct.
What I meant is that it does not always match the right hand operand in the test condition.
2011/8/25 Aric Stewart aric@codeweavers.com:
Ah i see what you are seeing.
I will see if that helps. However it is still strange that the tests are all working for me and when i submit tests but not when it is submitted as a patch.
-aric
On 8/25/11 7:47 AM, Hans Leidekker wrote:
On Thu, 2011-08-25 at 07:38 -0500, Aric Stewart wrote:
Yes, it depends on if it is a RTL or LTR string. That is correct.
What I meant is that it does not always match the right hand operand in the test condition.
toolchain differences ?
On 8/25/11 9:50 AM, Nicolas Le Cam wrote:
2011/8/25 Aric Stewartaric@codeweavers.com:
Ah i see what you are seeing.
I will see if that helps. However it is still strange that the tests are all working for me and when i submit tests but not when it is submitted as a patch.
-aric
On 8/25/11 7:47 AM, Hans Leidekker wrote:
On Thu, 2011-08-25 at 07:38 -0500, Aric Stewart wrote:
Yes, it depends on if it is a RTL or LTR string. That is correct.
What I meant is that it does not always match the right hand operand in the test condition.
toolchain differences ?
Yes, My compiler had uninitialized memory set to 0 (the value I was testing for) and the testbot did not.
-aric
-
Aric Stewart -
Hans Leidekker -
Hans Leidekker -
Marvin -
Nicolas Le Cam