You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@harmony.apache.org by Oliver Deakin <ol...@googlemail.com> on 2009/05/27 17:24:30 UTC
[M10] Testing
Hi all,
I've just run the classlib tests on Windows x86 on the latest code in
the federated build and I don't see anything catastrophic there. There's
1 failure in the portlib tests, in hymmap, which I think is a test case
off-by-one error. I've attached a patch [1] which fixes the test. Apart
from that we look clean.
Does anyone have any other test reports? Any other platforms? I'm
assuming we're taking r778555 as our testing base for M10.
Regards,
Oliver
[1]
Adding 2 to the path size for emptyFile and testFile is incorrect, as
pathLen already counts the rightmost DIR_SEPARATOR.
Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
===================================================================
--- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
779066)
+++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
@@ -126,8 +126,8 @@
execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the
rightmost slash character */
pathLen = strlen(execPath) - strlen(execName);
- emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
pathLen + strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the
extra slash and null terminator */
- testFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
pathLen + strlen("shared") + 2 + strlen("testFile"));
+ emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
pathLen + strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the
null terminator */
+ testFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
pathLen + strlen("shared") + 1 + strlen("testFile"));
strncpy(emptyFile, execPath, pathLen);
strcat(emptyFile, "shared");
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Nathan Beyer <nd...@apache.org>.
On Thu, May 28, 2009 at 8:28 AM, Oliver Deakin
<ol...@googlemail.com> wrote:
> Nathan Beyer wrote:
>>
>> I'd like to apply the final patch attached to
>> https://issues.apache.org/jira/browse/HARMONY-6074. This seems to
>> completely resolve the Maven test run issue.
>>
>
> Ok - since it seems the original patch applied at the end of the last week
> does not completely fix the issue, and our aim was to get the Maven tests
> running on M10, I'm +1 to this commit as well. I've applied the patch and
> run the luni tests and they all pass, so please go ahead and commit the
> patch as soon as possible so we can restart the testing process.
Applied.
>
>> As for testing, I'm not seeing the hymmap issue on Windows Vista x86.
>> However, the tests pass with and without this patch.
>>
>
> I've had the tests pass in the past as well, but there is definitely an
> error in the test case - what I see is that after the strncpy() on line 132
> of the hymmap test case we get an extra character at the end of the path.
> From looking at the msdn pages it looks like strncpy() does not add a null
> character to the end of the copied string if you are not copying the entire
> source string, which is the case here. Now I have had a closer look I think
> the +2 is actually correct, because we add _another_ slash character into
> the path further down, and in fact we simply need to make sure the string is
> null terminated after the strncpy() call so the strcat() call starts in the
> right place [1]. This patch makes the test behave as expected for me. As I
> said in the other branch of this thread, I don't feel this needs to be fixed
> for M10 so Ill hold off committing it for now.
>
> Regards,
> Oliver
This sounds good to me.
>
> [1]
> Index: hymmap.c
> ===================================================================
> --- hymmap.c (revision 779066)
> +++ hymmap.c (working copy)
> @@ -130,6 +130,7 @@
> testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 2 + strlen("testFile"));
>
> strncpy(emptyFile, execPath, pathLen);
> + emptyFile[pathLen] = '\0';
> strcat(emptyFile, "shared");
> strcat(emptyFile, DIR_SEPARATOR_STR);
> strcpy(testFile, emptyFile);
>
>> -Nathan
>>
>> On Wed, May 27, 2009 at 10:24 AM, Oliver Deakin
>> <ol...@googlemail.com> wrote:
>>
>>>
>>> Hi all,
>>>
>>> I've just run the classlib tests on Windows x86 on the latest code in the
>>> federated build and I don't see anything catastrophic there. There's 1
>>> failure in the portlib tests, in hymmap, which I think is a test case
>>> off-by-one error. I've attached a patch [1] which fixes the test. Apart
>>> from
>>> that we look clean.
>>>
>>> Does anyone have any other test reports? Any other platforms? I'm
>>> assuming
>>> we're taking r778555 as our testing base for M10.
>>>
>>> Regards,
>>> Oliver
>>>
>>> [1]
>>> Adding 2 to the path size for emptyFile and testFile is incorrect, as
>>> pathLen already counts the rightmost DIR_SEPARATOR.
>>>
>>> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
>>> ===================================================================
>>> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
>>> 779066)
>>> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working
>>> copy)
>>> @@ -126,8 +126,8 @@
>>> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the rightmost
>>> slash character */
>>> pathLen = strlen(execPath) - strlen(execName);
>>>
>>> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
>>> pathLen +
>>> strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the extra slash
>>> and
>>> null terminator */
>>> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen
>>> +
>>> strlen("shared") + 2 + strlen("testFile"));
>>> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
>>> pathLen +
>>> strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the null
>>> terminator
>>> */
>>> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen
>>> +
>>> strlen("shared") + 1 + strlen("testFile"));
>>>
>>> strncpy(emptyFile, execPath, pathLen);
>>> strcat(emptyFile, "shared");
>>>
>>> --
>>> Oliver Deakin
>>> Unless stated otherwise above:
>>> IBM United Kingdom Limited - Registered in England and Wales with number
>>> 741598. Registered office: PO Box 41, North Harbour, Portsmouth,
>>> Hampshire
>>> PO6 3AU
>>>
>>>
>>>
>>
>>
>
> --
> Oliver Deakin
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with number
> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
> PO6 3AU
>
>
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Nathan Beyer wrote:
> I'd like to apply the final patch attached to
> https://issues.apache.org/jira/browse/HARMONY-6074. This seems to
> completely resolve the Maven test run issue.
>
Ok - since it seems the original patch applied at the end of the last
week does not completely fix the issue, and our aim was to get the Maven
tests running on M10, I'm +1 to this commit as well. I've applied the
patch and run the luni tests and they all pass, so please go ahead and
commit the patch as soon as possible so we can restart the testing process.
> As for testing, I'm not seeing the hymmap issue on Windows Vista x86.
> However, the tests pass with and without this patch.
>
I've had the tests pass in the past as well, but there is definitely an
error in the test case - what I see is that after the strncpy() on line
132 of the hymmap test case we get an extra character at the end of the
path. From looking at the msdn pages it looks like strncpy() does not
add a null character to the end of the copied string if you are not
copying the entire source string, which is the case here. Now I have had
a closer look I think the +2 is actually correct, because we add
_another_ slash character into the path further down, and in fact we
simply need to make sure the string is null terminated after the
strncpy() call so the strcat() call starts in the right place [1]. This
patch makes the test behave as expected for me. As I said in the other
branch of this thread, I don't feel this needs to be fixed for M10 so
Ill hold off committing it for now.
Regards,
Oliver
[1]
Index: hymmap.c
===================================================================
--- hymmap.c (revision 779066)
+++ hymmap.c (working copy)
@@ -130,6 +130,7 @@
testFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
pathLen + strlen("shared") + 2 + strlen("testFile"));
strncpy(emptyFile, execPath, pathLen);
+ emptyFile[pathLen] = '\0';
strcat(emptyFile, "shared");
strcat(emptyFile, DIR_SEPARATOR_STR);
strcpy(testFile, emptyFile);
> -Nathan
>
> On Wed, May 27, 2009 at 10:24 AM, Oliver Deakin
> <ol...@googlemail.com> wrote:
>
>> Hi all,
>>
>> I've just run the classlib tests on Windows x86 on the latest code in the
>> federated build and I don't see anything catastrophic there. There's 1
>> failure in the portlib tests, in hymmap, which I think is a test case
>> off-by-one error. I've attached a patch [1] which fixes the test. Apart from
>> that we look clean.
>>
>> Does anyone have any other test reports? Any other platforms? I'm assuming
>> we're taking r778555 as our testing base for M10.
>>
>> Regards,
>> Oliver
>>
>> [1]
>> Adding 2 to the path size for emptyFile and testFile is incorrect, as
>> pathLen already counts the rightmost DIR_SEPARATOR.
>>
>> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
>> ===================================================================
>> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
>> 779066)
>> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
>> @@ -126,8 +126,8 @@
>> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the rightmost
>> slash character */
>> pathLen = strlen(execPath) - strlen(execName);
>>
>> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the extra slash and
>> null terminator */
>> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 2 + strlen("testFile"));
>> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the null terminator
>> */
>> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 1 + strlen("testFile"));
>>
>> strncpy(emptyFile, execPath, pathLen);
>> strcat(emptyFile, "shared");
>>
>> --
>> Oliver Deakin
>> Unless stated otherwise above:
>> IBM United Kingdom Limited - Registered in England and Wales with number
>> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>> PO6 3AU
>>
>>
>>
>
>
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Sean Qiu <se...@gmail.com>.
+1
Best Regards
Sean, Xiao Xia Qiu
2009/5/28 Nathan Beyer <nb...@gmail.com>:
> I'd like to apply the final patch attached to
> https://issues.apache.org/jira/browse/HARMONY-6074. This seems to
> completely resolve the Maven test run issue.
>
> As for testing, I'm not seeing the hymmap issue on Windows Vista x86.
> However, the tests pass with and without this patch.
>
> -Nathan
>
> On Wed, May 27, 2009 at 10:24 AM, Oliver Deakin
> <ol...@googlemail.com> wrote:
>> Hi all,
>>
>> I've just run the classlib tests on Windows x86 on the latest code in the
>> federated build and I don't see anything catastrophic there. There's 1
>> failure in the portlib tests, in hymmap, which I think is a test case
>> off-by-one error. I've attached a patch [1] which fixes the test. Apart from
>> that we look clean.
>>
>> Does anyone have any other test reports? Any other platforms? I'm assuming
>> we're taking r778555 as our testing base for M10.
>>
>> Regards,
>> Oliver
>>
>> [1]
>> Adding 2 to the path size for emptyFile and testFile is incorrect, as
>> pathLen already counts the rightmost DIR_SEPARATOR.
>>
>> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
>> ===================================================================
>> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
>> 779066)
>> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
>> @@ -126,8 +126,8 @@
>> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the rightmost
>> slash character */
>> pathLen = strlen(execPath) - strlen(execName);
>>
>> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the extra slash and
>> null terminator */
>> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 2 + strlen("testFile"));
>> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the null terminator
>> */
>> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 1 + strlen("testFile"));
>>
>> strncpy(emptyFile, execPath, pathLen);
>> strcat(emptyFile, "shared");
>>
>> --
>> Oliver Deakin
>> Unless stated otherwise above:
>> IBM United Kingdom Limited - Registered in England and Wales with number
>> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>> PO6 3AU
>>
>>
>
Re: [M10] Testing
Posted by Nathan Beyer <nb...@gmail.com>.
I'd like to apply the final patch attached to
https://issues.apache.org/jira/browse/HARMONY-6074. This seems to
completely resolve the Maven test run issue.
As for testing, I'm not seeing the hymmap issue on Windows Vista x86.
However, the tests pass with and without this patch.
-Nathan
On Wed, May 27, 2009 at 10:24 AM, Oliver Deakin
<ol...@googlemail.com> wrote:
> Hi all,
>
> I've just run the classlib tests on Windows x86 on the latest code in the
> federated build and I don't see anything catastrophic there. There's 1
> failure in the portlib tests, in hymmap, which I think is a test case
> off-by-one error. I've attached a patch [1] which fixes the test. Apart from
> that we look clean.
>
> Does anyone have any other test reports? Any other platforms? I'm assuming
> we're taking r778555 as our testing base for M10.
>
> Regards,
> Oliver
>
> [1]
> Adding 2 to the path size for emptyFile and testFile is incorrect, as
> pathLen already counts the rightmost DIR_SEPARATOR.
>
> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
> ===================================================================
> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
> 779066)
> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
> @@ -126,8 +126,8 @@
> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the rightmost
> slash character */
> pathLen = strlen(execPath) - strlen(execName);
>
> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the extra slash and
> null terminator */
> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 2 + strlen("testFile"));
> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the null terminator
> */
> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 1 + strlen("testFile"));
>
> strncpy(emptyFile, execPath, pathLen);
> strcat(emptyFile, "shared");
>
> --
> Oliver Deakin
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with number
> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
> PO6 3AU
>
>
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Hi Sian,
I don't feel strongly either way on committing the fix to the test case.
Once the fix is applied, the test passes for me so I don't think there
is a bug in the mmap code we have to worry about. I'm happy to wait
until after M10 for this patch to be applied just to keep change in the
codebase to a minimum during code freeze.
Regards,
Oliver
Sian January wrote:
> I see the hymmap test failure too on Windows XP, but nothing else
> apart from a couple of AWT and Swing failures, which were also there
> in M9.
>
> Shall we leave your patch until after M10 as it's a testcase issue
> rather than a code issue?
>
>
> 2009/5/27 Oliver Deakin <ol...@googlemail.com>:
>
>> Hi all,
>>
>> I've just run the classlib tests on Windows x86 on the latest code in the
>> federated build and I don't see anything catastrophic there. There's 1
>> failure in the portlib tests, in hymmap, which I think is a test case
>> off-by-one error. I've attached a patch [1] which fixes the test. Apart from
>> that we look clean.
>>
>> Does anyone have any other test reports? Any other platforms? I'm assuming
>> we're taking r778555 as our testing base for M10.
>>
>> Regards,
>> Oliver
>>
>> [1]
>> Adding 2 to the path size for emptyFile and testFile is incorrect, as
>> pathLen already counts the rightmost DIR_SEPARATOR.
>>
>> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
>> ===================================================================
>> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
>> 779066)
>> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
>> @@ -126,8 +126,8 @@
>> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the rightmost
>> slash character */
>> pathLen = strlen(execPath) - strlen(execName);
>>
>> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the extra slash and
>> null terminator */
>> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 2 + strlen("testFile"));
>> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the null terminator
>> */
>> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
>> strlen("shared") + 1 + strlen("testFile"));
>>
>> strncpy(emptyFile, execPath, pathLen);
>> strcat(emptyFile, "shared");
>>
>> --
>> Oliver Deakin
>> Unless stated otherwise above:
>> IBM United Kingdom Limited - Registered in England and Wales with number
>> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>> PO6 3AU
>>
>>
>>
>
>
>
>
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Sian January <si...@googlemail.com>.
I see the hymmap test failure too on Windows XP, but nothing else
apart from a couple of AWT and Swing failures, which were also there
in M9.
Shall we leave your patch until after M10 as it's a testcase issue
rather than a code issue?
2009/5/27 Oliver Deakin <ol...@googlemail.com>:
> Hi all,
>
> I've just run the classlib tests on Windows x86 on the latest code in the
> federated build and I don't see anything catastrophic there. There's 1
> failure in the portlib tests, in hymmap, which I think is a test case
> off-by-one error. I've attached a patch [1] which fixes the test. Apart from
> that we look clean.
>
> Does anyone have any other test reports? Any other platforms? I'm assuming
> we're taking r778555 as our testing base for M10.
>
> Regards,
> Oliver
>
> [1]
> Adding 2 to the path size for emptyFile and testFile is incorrect, as
> pathLen already counts the rightmost DIR_SEPARATOR.
>
> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
> ===================================================================
> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
> 779066)
> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
> @@ -126,8 +126,8 @@
> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the rightmost
> slash character */
> pathLen = strlen(execPath) - strlen(execName);
>
> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the extra slash and
> null terminator */
> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 2 + strlen("testFile"));
> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the null terminator
> */
> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary, pathLen +
> strlen("shared") + 1 + strlen("testFile"));
>
> strncpy(emptyFile, execPath, pathLen);
> strcat(emptyFile, "shared");
>
> --
> Oliver Deakin
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with number
> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
> PO6 3AU
>
>
--
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Nathan Beyer wrote:
> I changed most of the issues from 5M10 to Unknown, as they were issues
> that existed with M9 and either didn't show up until code freeze or
> didn't get looked at.
>
> We just need to get in the practice of leaving issues unknown until
> they're fixed and then mark the fix version.
>
+1 on this, although if someone feels a bug is truly must-fix for a
particular milestone then I think they should target it as such.
Regards,
Oliver
> I would like to see the NOTICE and README issues fixed or commented
> on. I think Tim was working on one of them.
>
> -Nathan
>
> On Tue, Jun 2, 2009 at 8:08 AM, Oliver Deakin
> <ol...@googlemail.com> wrote:
>
>> Yep, you're totally correct - I'd missed those last issues. As you say, I
>> think HARMONY-6193 and HARMONY-6155 are must fix, but are not functional so
>> do not block testing.
>>
>> Does anyone think that any of the other issues in [1] are must-fix for M10?
>> My personal feeling is that they are not and can be moved to an M11 target,
>> but please speak up if you feel differently.
>>
>> Regards,
>> Oliver
>>
>> [1]
>> http://issues.apache.org/jira/secure/IssueNavigator.jspa?reset=true&mode=hide&sorter/order=DESC&sorter/field=priority&resolution=-1&pid=12310050&fixfor=12313869
>>
>> Tim Ellison wrote:
>>
>>> I still see a bunch of JIRA issues flagged as to be fixed in M10. I'm
>>> guessing these are mis-labeled except HARMONY-6193 and HARMONY-6155.
>>>
>>> Regards,
>>> Tim
>>>
>>> Oliver Deakin wrote:
>>>
>>>
>>>> Let's consider repo revision r780017 as M10 - I've just run all the
>>>> classlib tests on Windows XP x86 and they have all passed successfully
>>>> for me, so no adverse effects from the recent changes.
>>>>
>>>> Does anyone have any more test results for any other platforms? It would
>>>> be nice to have at least Linux x86 (and perhaps linux x86_64) if
>>>> possible...
>>>>
>>>> Im going to try and do some app testing as best I can this week - if
>>>> anyone gets any time to do this, please post your results here.
>>>>
>>>> Regards,
>>>> Oliver
>>>>
>>>> Sian January wrote:
>>>>
>>>>
>>>>> Tests are still looking good for me on Windows after Mark's commit.
>>>>>
>>>>>
>>>>> 2009/5/29 Mark Hindess <ma...@googlemail.com>:
>>>>>
>>>>>
>>>>>> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing
>>>>>> in
>>>>>> the sun.)
>>>>>>
>>>>>> -Mark.
>>>>>>
>>>>>> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>>>>>
>>>>>>
>>>>>>> Looks like the patch has 2 committers supporting - I'm also +1 to
>>>>>>> fixing
>>>>>>> the test crashes and if the patch for HARMONY-6132 does this without
>>>>>>> affecting any of the other tests, then Im also +1 to the patch. Please
>>>>>>> go ahead and commit it as soon as possible so we can begin final
>>>>>>> testing.
>>>>>>>
>>>>>>> Regards,
>>>>>>> Oliver
>>>>>>>
>>>>>>> Alexey Varlamov wrote:
>>>>>>>
>>>>>>>
>>>>>>>> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>>>>>>>> [skipped]
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> I also had two test crashes:
>>>>>>>>>
>>>>>>>>> org.apache.harmony.luni.tests.java.net.SocketTest
>>>>>>>>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>>>>>>>>
>>>>>>>>> which were both fixed by Ilya's patch in HARMONY-6132.
>>>>>>>>>
>>>>>>>>> I think it would be good to apply the patch from HARMONY-6132
>>>>>>>>> before M10.
>>>>>>>>> Does anyone else support/oppose this?
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>> The fix looks reasonable and the issue is important enough to fix
>>>>>>>> it in M10
>>>>>>>>
>>>>>>>>
>>>>>>> .
>>>>>>>
>>>>>>>
>>>>>>>> --
>>>>>>>> Alexey
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> Regards,
>>>>>>>>> Mark.
>>>>>>>>>
>>>>>>>>>
>>>>>>> --
>>>>>>> Oliver Deakin
>>>>>>> Unless stated otherwise above:
>>>>>>> IBM United Kingdom Limited - Registered in England and Wales with
>>>>>>> number 7415
>>>>>>> 98.
>>>>>>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>>>>>>> PO6 3AU
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>
>> --
>> Oliver Deakin
>> Unless stated otherwise above:
>> IBM United Kingdom Limited - Registered in England and Wales with number
>> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>> PO6 3AU
>>
>>
>>
>
>
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Sian January <si...@googlemail.com>.
I think that looks fine in general, but it might be better not to
mention Java 5 specifically since we are also working towards Java 6
and our Java 6 snapshots would presumably contain this readme after
the next merge.
2009/6/3 Mark Hindess <ma...@googlemail.com>:
>
> In message <3b...@mail.gmail.com>,
> Nathan Beyer writes:
>>
>> I changed most of the issues from 5M10 to Unknown, as they were issues
>> that existed with M9 and either didn't show up until code freeze or
>> didn't get looked at.
>>
>> We just need to get in the practice of leaving issues unknown until
>> they're fixed and then mark the fix version.
>>
>> I would like to see the NOTICE and README issues fixed or commented
>> on. I think Tim was working on one of them.
>
> I think we need a top-level README file with the crucial export notice.
> This README can then be included in all the jars. We don't seem to have
> a top-level README anywhere but perhaps something like the appended text
> would do for now?
>
> I think that would be acceptable for M10 but we can always refine the
> top-level README later and consider whether more specific README files
> might be appropriate for individual jars.
>
> I'd appreciate comments and/or permission to commit the README and
> necessary jar changes (which I'm preparing a patch for)?
>
> Regards,
> Mark.
>
> --------------------------------------------------------------------------
>
> Apache Harmony - Open Source Java SE
> http://harmony.apache.org/
>
> Apache Harmony is the Java SE project of the Apache Software Foundation.
>
> The aim of the project is to produce a large and healthy community of
> those interested in runtime platforms tasked with creation of:
>
> * A compatible, independent implementation of the Java SE 5 JDK under
> the Apache License v2
>
> * A community-developed modular runtime (VM and class library)
> architecture.
>
> For more information about using Apache Harmony see:
>
> http://harmony.apache.org/quickhelp_users.html
>
>
> Export Notice
> -------------
>
> This distribution includes cryptographic software. The country in
> which you currently reside may have restrictions on the import,
> possession, use, and/or re-export to another country, of
> encryption software. BEFORE using any encryption software, please
> check your country's laws, regulations and policies concerning the
> import, possession, or use, and re-export of encryption software, to
> see if this is permitted. See http://www.wassenaar.org/ for more
> information.
>
> The U.S. Government Department of Commerce, Bureau of Industry and
> Security (BIS), has classified this software as Export Commodity
> Control Number (ECCN) 5D002.C.1, which includes information security
> software using or performing cryptographic functions with asymmetric
> algorithms. The form and manner of this Apache Software Foundation
> distribution makes it eligible for export under the License Exception
> ENC Technology Software Unrestricted (TSU) exception (see the BIS
> Export Administration Regulations, Section 740.13) for both object
> code and source code.
>
> The following provides more details on the included cryptographic
> software:
>
> Apache Harmony contains code that is specifically designed to enable
> cryptography. In particular Apache Harmony contains an implementation
> the Java cryptographic extensions. In addition, binary distributions
> of Apache Harmony may contain cryptographic functionality provided by
> The Legion of the Bouncy Castle (http://www.bouncycastle.org).
>
>
>
--
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Tim Ellison <t....@gmail.com>.
We are not obliged to include the readme in all JAR files, just once at
the top level of the distribution. However, we /do/ need to include the
license and notice files in each JARs.
Regards,
Tim
Mark Hindess wrote:
> In message <3b...@mail.gmail.com>,
> Nathan Beyer writes:
>> I changed most of the issues from 5M10 to Unknown, as they were issues
>> that existed with M9 and either didn't show up until code freeze or
>> didn't get looked at.
>>
>> We just need to get in the practice of leaving issues unknown until
>> they're fixed and then mark the fix version.
>>
>> I would like to see the NOTICE and README issues fixed or commented
>> on. I think Tim was working on one of them.
>
> I think we need a top-level README file with the crucial export notice.
> This README can then be included in all the jars. We don't seem to have
> a top-level README anywhere but perhaps something like the appended text
> would do for now?
>
> I think that would be acceptable for M10 but we can always refine the
> top-level README later and consider whether more specific README files
> might be appropriate for individual jars.
>
> I'd appreciate comments and/or permission to commit the README and
> necessary jar changes (which I'm preparing a patch for)?
>
> Regards,
> Mark.
>
> --------------------------------------------------------------------------
>
> Apache Harmony - Open Source Java SE
> http://harmony.apache.org/
>
> Apache Harmony is the Java SE project of the Apache Software Foundation.
>
> The aim of the project is to produce a large and healthy community of
> those interested in runtime platforms tasked with creation of:
>
> * A compatible, independent implementation of the Java SE 5 JDK under
> the Apache License v2
>
> * A community-developed modular runtime (VM and class library)
> architecture.
>
> For more information about using Apache Harmony see:
>
> http://harmony.apache.org/quickhelp_users.html
>
>
> Export Notice
> -------------
>
> This distribution includes cryptographic software. The country in
> which you currently reside may have restrictions on the import,
> possession, use, and/or re-export to another country, of
> encryption software. BEFORE using any encryption software, please
> check your country's laws, regulations and policies concerning the
> import, possession, or use, and re-export of encryption software, to
> see if this is permitted. See http://www.wassenaar.org/ for more
> information.
>
> The U.S. Government Department of Commerce, Bureau of Industry and
> Security (BIS), has classified this software as Export Commodity
> Control Number (ECCN) 5D002.C.1, which includes information security
> software using or performing cryptographic functions with asymmetric
> algorithms. The form and manner of this Apache Software Foundation
> distribution makes it eligible for export under the License Exception
> ENC Technology Software Unrestricted (TSU) exception (see the BIS
> Export Administration Regulations, Section 740.13) for both object
> code and source code.
>
> The following provides more details on the included cryptographic
> software:
>
> Apache Harmony contains code that is specifically designed to enable
> cryptography. In particular Apache Harmony contains an implementation
> the Java cryptographic extensions. In addition, binary distributions
> of Apache Harmony may contain cryptographic functionality provided by
> The Legion of the Bouncy Castle (http://www.bouncycastle.org).
>
>
>
Re: [M10] Testing
Posted by Mark Hindess <ma...@googlemail.com>.
In message <3b...@mail.gmail.com>,
Nathan Beyer writes:
>
> I changed most of the issues from 5M10 to Unknown, as they were issues
> that existed with M9 and either didn't show up until code freeze or
> didn't get looked at.
>
> We just need to get in the practice of leaving issues unknown until
> they're fixed and then mark the fix version.
>
> I would like to see the NOTICE and README issues fixed or commented
> on. I think Tim was working on one of them.
I think we need a top-level README file with the crucial export notice.
This README can then be included in all the jars. We don't seem to have
a top-level README anywhere but perhaps something like the appended text
would do for now?
I think that would be acceptable for M10 but we can always refine the
top-level README later and consider whether more specific README files
might be appropriate for individual jars.
I'd appreciate comments and/or permission to commit the README and
necessary jar changes (which I'm preparing a patch for)?
Regards,
Mark.
--------------------------------------------------------------------------
Apache Harmony - Open Source Java SE
http://harmony.apache.org/
Apache Harmony is the Java SE project of the Apache Software Foundation.
The aim of the project is to produce a large and healthy community of
those interested in runtime platforms tasked with creation of:
* A compatible, independent implementation of the Java SE 5 JDK under
the Apache License v2
* A community-developed modular runtime (VM and class library)
architecture.
For more information about using Apache Harmony see:
http://harmony.apache.org/quickhelp_users.html
Export Notice
-------------
This distribution includes cryptographic software. The country in
which you currently reside may have restrictions on the import,
possession, use, and/or re-export to another country, of
encryption software. BEFORE using any encryption software, please
check your country's laws, regulations and policies concerning the
import, possession, or use, and re-export of encryption software, to
see if this is permitted. See http://www.wassenaar.org/ for more
information.
The U.S. Government Department of Commerce, Bureau of Industry and
Security (BIS), has classified this software as Export Commodity
Control Number (ECCN) 5D002.C.1, which includes information security
software using or performing cryptographic functions with asymmetric
algorithms. The form and manner of this Apache Software Foundation
distribution makes it eligible for export under the License Exception
ENC Technology Software Unrestricted (TSU) exception (see the BIS
Export Administration Regulations, Section 740.13) for both object
code and source code.
The following provides more details on the included cryptographic
software:
Apache Harmony contains code that is specifically designed to enable
cryptography. In particular Apache Harmony contains an implementation
the Java cryptographic extensions. In addition, binary distributions
of Apache Harmony may contain cryptographic functionality provided by
The Legion of the Bouncy Castle (http://www.bouncycastle.org).
Re: [M10] Testing
Posted by Nathan Beyer <nd...@apache.org>.
I changed most of the issues from 5M10 to Unknown, as they were issues
that existed with M9 and either didn't show up until code freeze or
didn't get looked at.
We just need to get in the practice of leaving issues unknown until
they're fixed and then mark the fix version.
I would like to see the NOTICE and README issues fixed or commented
on. I think Tim was working on one of them.
-Nathan
On Tue, Jun 2, 2009 at 8:08 AM, Oliver Deakin
<ol...@googlemail.com> wrote:
> Yep, you're totally correct - I'd missed those last issues. As you say, I
> think HARMONY-6193 and HARMONY-6155 are must fix, but are not functional so
> do not block testing.
>
> Does anyone think that any of the other issues in [1] are must-fix for M10?
> My personal feeling is that they are not and can be moved to an M11 target,
> but please speak up if you feel differently.
>
> Regards,
> Oliver
>
> [1]
> http://issues.apache.org/jira/secure/IssueNavigator.jspa?reset=true&mode=hide&sorter/order=DESC&sorter/field=priority&resolution=-1&pid=12310050&fixfor=12313869
>
> Tim Ellison wrote:
>>
>> I still see a bunch of JIRA issues flagged as to be fixed in M10. I'm
>> guessing these are mis-labeled except HARMONY-6193 and HARMONY-6155.
>>
>> Regards,
>> Tim
>>
>> Oliver Deakin wrote:
>>
>>>
>>> Let's consider repo revision r780017 as M10 - I've just run all the
>>> classlib tests on Windows XP x86 and they have all passed successfully
>>> for me, so no adverse effects from the recent changes.
>>>
>>> Does anyone have any more test results for any other platforms? It would
>>> be nice to have at least Linux x86 (and perhaps linux x86_64) if
>>> possible...
>>>
>>> Im going to try and do some app testing as best I can this week - if
>>> anyone gets any time to do this, please post your results here.
>>>
>>> Regards,
>>> Oliver
>>>
>>> Sian January wrote:
>>>
>>>>
>>>> Tests are still looking good for me on Windows after Mark's commit.
>>>>
>>>>
>>>> 2009/5/29 Mark Hindess <ma...@googlemail.com>:
>>>>
>>>>>
>>>>> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing
>>>>> in
>>>>> the sun.)
>>>>>
>>>>> -Mark.
>>>>>
>>>>> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>>>>
>>>>>>
>>>>>> Looks like the patch has 2 committers supporting - I'm also +1 to
>>>>>> fixing
>>>>>> the test crashes and if the patch for HARMONY-6132 does this without
>>>>>> affecting any of the other tests, then Im also +1 to the patch. Please
>>>>>> go ahead and commit it as soon as possible so we can begin final
>>>>>> testing.
>>>>>>
>>>>>> Regards,
>>>>>> Oliver
>>>>>>
>>>>>> Alexey Varlamov wrote:
>>>>>>
>>>>>>>
>>>>>>> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>>>>>>> [skipped]
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> I also had two test crashes:
>>>>>>>>
>>>>>>>> org.apache.harmony.luni.tests.java.net.SocketTest
>>>>>>>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>>>>>>>
>>>>>>>> which were both fixed by Ilya's patch in HARMONY-6132.
>>>>>>>>
>>>>>>>> I think it would be good to apply the patch from HARMONY-6132
>>>>>>>> before M10.
>>>>>>>> Does anyone else support/oppose this?
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> The fix looks reasonable and the issue is important enough to fix
>>>>>>> it in M10
>>>>>>>
>>>>>>
>>>>>> .
>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Alexey
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>> Mark.
>>>>>>>>
>>>>>>
>>>>>> --
>>>>>> Oliver Deakin
>>>>>> Unless stated otherwise above:
>>>>>> IBM United Kingdom Limited - Registered in England and Wales with
>>>>>> number 7415
>>>>>> 98.
>>>>>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>>>>>> PO6 3AU
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>
>>
>
> --
> Oliver Deakin
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with number
> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
> PO6 3AU
>
>
Re: [M10] Testing
Posted by Sian January <si...@googlemail.com>.
2009/6/2 Oliver Deakin <ol...@googlemail.com>:
> Yep, you're totally correct - I'd missed those last issues. As you say, I
> think HARMONY-6193 and HARMONY-6155 are must fix, but are not functional so
> do not block testing.
What is the status of these two issues? For HARMONY-6155 it seems
like the LICENSE file is up to date, but there may still be some
notices missing from the NOTICE file depending on what needs to go in
there. Is that right?
>
> Does anyone think that any of the other issues in [1] are must-fix for M10?
> My personal feeling is that they are not and can be moved to an M11 target,
> but please speak up if you feel differently.
Looking at the dates these other issues were only raised in the last
week so I agree that it's probably a mistake. Perhaps Kevin can
confirm this?
>
> Regards,
> Oliver
>
> [1]
> http://issues.apache.org/jira/secure/IssueNavigator.jspa?reset=true&mode=hide&sorter/order=DESC&sorter/field=priority&resolution=-1&pid=12310050&fixfor=12313869
>
> Tim Ellison wrote:
>>
>> I still see a bunch of JIRA issues flagged as to be fixed in M10. I'm
>> guessing these are mis-labeled except HARMONY-6193 and HARMONY-6155.
>>
>> Regards,
>> Tim
>>
>> Oliver Deakin wrote:
>>
>>>
>>> Let's consider repo revision r780017 as M10 - I've just run all the
>>> classlib tests on Windows XP x86 and they have all passed successfully
>>> for me, so no adverse effects from the recent changes.
>>>
>>> Does anyone have any more test results for any other platforms? It would
>>> be nice to have at least Linux x86 (and perhaps linux x86_64) if
>>> possible...
>>>
>>> Im going to try and do some app testing as best I can this week - if
>>> anyone gets any time to do this, please post your results here.
>>>
>>> Regards,
>>> Oliver
>>>
>>> Sian January wrote:
>>>
>>>>
>>>> Tests are still looking good for me on Windows after Mark's commit.
>>>>
>>>>
>>>> 2009/5/29 Mark Hindess <ma...@googlemail.com>:
>>>>
>>>>>
>>>>> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing
>>>>> in
>>>>> the sun.)
>>>>>
>>>>> -Mark.
>>>>>
>>>>> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>>>>
>>>>>>
>>>>>> Looks like the patch has 2 committers supporting - I'm also +1 to
>>>>>> fixing
>>>>>> the test crashes and if the patch for HARMONY-6132 does this without
>>>>>> affecting any of the other tests, then Im also +1 to the patch. Please
>>>>>> go ahead and commit it as soon as possible so we can begin final
>>>>>> testing.
>>>>>>
>>>>>> Regards,
>>>>>> Oliver
>>>>>>
>>>>>> Alexey Varlamov wrote:
>>>>>>
>>>>>>>
>>>>>>> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>>>>>>> [skipped]
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> I also had two test crashes:
>>>>>>>>
>>>>>>>> org.apache.harmony.luni.tests.java.net.SocketTest
>>>>>>>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>>>>>>>
>>>>>>>> which were both fixed by Ilya's patch in HARMONY-6132.
>>>>>>>>
>>>>>>>> I think it would be good to apply the patch from HARMONY-6132
>>>>>>>> before M10.
>>>>>>>> Does anyone else support/oppose this?
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> The fix looks reasonable and the issue is important enough to fix
>>>>>>> it in M10
>>>>>>>
>>>>>>
>>>>>> .
>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Alexey
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>> Mark.
>>>>>>>>
>>>>>>
>>>>>> --
>>>>>> Oliver Deakin
>>>>>> Unless stated otherwise above:
>>>>>> IBM United Kingdom Limited - Registered in England and Wales with
>>>>>> number 7415
>>>>>> 98.
>>>>>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>>>>>> PO6 3AU
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>
>>
>
> --
> Oliver Deakin
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with number
> 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
> PO6 3AU
>
>
--
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Yep, you're totally correct - I'd missed those last issues. As you say,
I think HARMONY-6193 and HARMONY-6155 are must fix, but are not
functional so do not block testing.
Does anyone think that any of the other issues in [1] are must-fix for
M10? My personal feeling is that they are not and can be moved to an M11
target, but please speak up if you feel differently.
Regards,
Oliver
[1]
http://issues.apache.org/jira/secure/IssueNavigator.jspa?reset=true&mode=hide&sorter/order=DESC&sorter/field=priority&resolution=-1&pid=12310050&fixfor=12313869
Tim Ellison wrote:
> I still see a bunch of JIRA issues flagged as to be fixed in M10. I'm
> guessing these are mis-labeled except HARMONY-6193 and HARMONY-6155.
>
> Regards,
> Tim
>
> Oliver Deakin wrote:
>
>> Let's consider repo revision r780017 as M10 - I've just run all the
>> classlib tests on Windows XP x86 and they have all passed successfully
>> for me, so no adverse effects from the recent changes.
>>
>> Does anyone have any more test results for any other platforms? It would
>> be nice to have at least Linux x86 (and perhaps linux x86_64) if
>> possible...
>>
>> Im going to try and do some app testing as best I can this week - if
>> anyone gets any time to do this, please post your results here.
>>
>> Regards,
>> Oliver
>>
>> Sian January wrote:
>>
>>> Tests are still looking good for me on Windows after Mark's commit.
>>>
>>>
>>> 2009/5/29 Mark Hindess <ma...@googlemail.com>:
>>>
>>>
>>>> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing in
>>>> the sun.)
>>>>
>>>> -Mark.
>>>>
>>>> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>>>
>>>>
>>>>> Looks like the patch has 2 committers supporting - I'm also +1 to
>>>>> fixing
>>>>> the test crashes and if the patch for HARMONY-6132 does this without
>>>>> affecting any of the other tests, then Im also +1 to the patch. Please
>>>>> go ahead and commit it as soon as possible so we can begin final
>>>>> testing.
>>>>>
>>>>> Regards,
>>>>> Oliver
>>>>>
>>>>> Alexey Varlamov wrote:
>>>>>
>>>>>
>>>>>> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>>>>>> [skipped]
>>>>>>
>>>>>>
>>>>>>
>>>>>>> I also had two test crashes:
>>>>>>>
>>>>>>> org.apache.harmony.luni.tests.java.net.SocketTest
>>>>>>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>>>>>>
>>>>>>> which were both fixed by Ilya's patch in HARMONY-6132.
>>>>>>>
>>>>>>> I think it would be good to apply the patch from HARMONY-6132
>>>>>>> before M10.
>>>>>>> Does anyone else support/oppose this?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> The fix looks reasonable and the issue is important enough to fix
>>>>>> it in M10
>>>>>>
>>>>>>
>>>>> .
>>>>>
>>>>>
>>>>>> --
>>>>>> Alexey
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Regards,
>>>>>>> Mark.
>>>>>>>
>>>>>>>
>>>>> --
>>>>> Oliver Deakin
>>>>> Unless stated otherwise above:
>>>>> IBM United Kingdom Limited - Registered in England and Wales with
>>>>> number 7415
>>>>> 98.
>>>>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>>>>> PO6 3AU
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>>
>
>
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Tim Ellison <t....@gmail.com>.
I still see a bunch of JIRA issues flagged as to be fixed in M10. I'm
guessing these are mis-labeled except HARMONY-6193 and HARMONY-6155.
Regards,
Tim
Oliver Deakin wrote:
> Let's consider repo revision r780017 as M10 - I've just run all the
> classlib tests on Windows XP x86 and they have all passed successfully
> for me, so no adverse effects from the recent changes.
>
> Does anyone have any more test results for any other platforms? It would
> be nice to have at least Linux x86 (and perhaps linux x86_64) if
> possible...
>
> Im going to try and do some app testing as best I can this week - if
> anyone gets any time to do this, please post your results here.
>
> Regards,
> Oliver
>
> Sian January wrote:
>> Tests are still looking good for me on Windows after Mark's commit.
>>
>>
>> 2009/5/29 Mark Hindess <ma...@googlemail.com>:
>>
>>> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing in
>>> the sun.)
>>>
>>> -Mark.
>>>
>>> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>>
>>>> Looks like the patch has 2 committers supporting - I'm also +1 to
>>>> fixing
>>>> the test crashes and if the patch for HARMONY-6132 does this without
>>>> affecting any of the other tests, then Im also +1 to the patch. Please
>>>> go ahead and commit it as soon as possible so we can begin final
>>>> testing.
>>>>
>>>> Regards,
>>>> Oliver
>>>>
>>>> Alexey Varlamov wrote:
>>>>
>>>>> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>>>>> [skipped]
>>>>>
>>>>>
>>>>>> I also had two test crashes:
>>>>>>
>>>>>> org.apache.harmony.luni.tests.java.net.SocketTest
>>>>>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>>>>>
>>>>>> which were both fixed by Ilya's patch in HARMONY-6132.
>>>>>>
>>>>>> I think it would be good to apply the patch from HARMONY-6132
>>>>>> before M10.
>>>>>> Does anyone else support/oppose this?
>>>>>>
>>>>>>
>>>>> The fix looks reasonable and the issue is important enough to fix
>>>>> it in M10
>>>>>
>>>> .
>>>>
>>>>> --
>>>>> Alexey
>>>>>
>>>>>
>>>>>
>>>>>> Regards,
>>>>>> Mark.
>>>>>>
>>>> --
>>>> Oliver Deakin
>>>> Unless stated otherwise above:
>>>> IBM United Kingdom Limited - Registered in England and Wales with
>>>> number 7415
>>>> 98.
>>>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
>>>> PO6 3AU
>>>>
>>>>
>>>
>>>
>>
>>
>>
>>
>
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Let's consider repo revision r780017 as M10 - I've just run all the
classlib tests on Windows XP x86 and they have all passed successfully
for me, so no adverse effects from the recent changes.
Does anyone have any more test results for any other platforms? It would
be nice to have at least Linux x86 (and perhaps linux x86_64) if possible...
Im going to try and do some app testing as best I can this week - if
anyone gets any time to do this, please post your results here.
Regards,
Oliver
Sian January wrote:
> Tests are still looking good for me on Windows after Mark's commit.
>
>
> 2009/5/29 Mark Hindess <ma...@googlemail.com>:
>
>> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing in
>> the sun.)
>>
>> -Mark.
>>
>> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>
>>> Looks like the patch has 2 committers supporting - I'm also +1 to fixing
>>> the test crashes and if the patch for HARMONY-6132 does this without
>>> affecting any of the other tests, then Im also +1 to the patch. Please
>>> go ahead and commit it as soon as possible so we can begin final testing.
>>>
>>> Regards,
>>> Oliver
>>>
>>> Alexey Varlamov wrote:
>>>
>>>> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>>>> [skipped]
>>>>
>>>>
>>>>> I also had two test crashes:
>>>>>
>>>>> org.apache.harmony.luni.tests.java.net.SocketTest
>>>>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>>>>
>>>>> which were both fixed by Ilya's patch in HARMONY-6132.
>>>>>
>>>>> I think it would be good to apply the patch from HARMONY-6132 before M10.
>>>>> Does anyone else support/oppose this?
>>>>>
>>>>>
>>>> The fix looks reasonable and the issue is important enough to fix it in M10
>>>>
>>> .
>>>
>>>> --
>>>> Alexey
>>>>
>>>>
>>>>
>>>>> Regards,
>>>>> Mark.
>>>>>
>>> --
>>> Oliver Deakin
>>> Unless stated otherwise above:
>>> IBM United Kingdom Limited - Registered in England and Wales with number 7415
>>> 98.
>>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
>>>
>>>
>>
>>
>
>
>
>
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Sian January <si...@googlemail.com>.
Tests are still looking good for me on Windows after Mark's commit.
2009/5/29 Mark Hindess <ma...@googlemail.com>:
>
> Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing in
> the sun.)
>
> -Mark.
>
> In message <4A...@googlemail.com>, Oliver Deakin writes:
>>
>> Looks like the patch has 2 committers supporting - I'm also +1 to fixing
>> the test crashes and if the patch for HARMONY-6132 does this without
>> affecting any of the other tests, then Im also +1 to the patch. Please
>> go ahead and commit it as soon as possible so we can begin final testing.
>>
>> Regards,
>> Oliver
>>
>> Alexey Varlamov wrote:
>> > 2009/5/28, Mark Hindess <ma...@googlemail.com>:
>> > [skipped]
>> >
>> >> I also had two test crashes:
>> >>
>> >> org.apache.harmony.luni.tests.java.net.SocketTest
>> >> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>> >>
>> >> which were both fixed by Ilya's patch in HARMONY-6132.
>> >>
>> >> I think it would be good to apply the patch from HARMONY-6132 before M10.
>> >> Does anyone else support/oppose this?
>> >>
>> >
>> > The fix looks reasonable and the issue is important enough to fix it in M10
>> .
>> > --
>> > Alexey
>> >
>> >
>> >> Regards,
>> >> Mark.
>>
>> --
>> Oliver Deakin
>> Unless stated otherwise above:
>> IBM United Kingdom Limited - Registered in England and Wales with number 7415
>> 98.
>> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
>>
>
>
>
--
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Mark Hindess <ma...@googlemail.com>.
Applied at r780017. (Sorry I was a bit slow; I was too busy relaxing in
the sun.)
-Mark.
In message <4A...@googlemail.com>, Oliver Deakin writes:
>
> Looks like the patch has 2 committers supporting - I'm also +1 to fixing
> the test crashes and if the patch for HARMONY-6132 does this without
> affecting any of the other tests, then Im also +1 to the patch. Please
> go ahead and commit it as soon as possible so we can begin final testing.
>
> Regards,
> Oliver
>
> Alexey Varlamov wrote:
> > 2009/5/28, Mark Hindess <ma...@googlemail.com>:
> > [skipped]
> >
> >> I also had two test crashes:
> >>
> >> org.apache.harmony.luni.tests.java.net.SocketTest
> >> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
> >>
> >> which were both fixed by Ilya's patch in HARMONY-6132.
> >>
> >> I think it would be good to apply the patch from HARMONY-6132 before M10.
> >> Does anyone else support/oppose this?
> >>
> >
> > The fix looks reasonable and the issue is important enough to fix it in M10
> .
> > --
> > Alexey
> >
> >
> >> Regards,
> >> Mark.
>
> --
> Oliver Deakin
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with number 7415
> 98.
> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
>
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Looks like the patch has 2 committers supporting - I'm also +1 to fixing
the test crashes and if the patch for HARMONY-6132 does this without
affecting any of the other tests, then Im also +1 to the patch. Please
go ahead and commit it as soon as possible so we can begin final testing.
Regards,
Oliver
Alexey Varlamov wrote:
> 2009/5/28, Mark Hindess <ma...@googlemail.com>:
> [skipped]
>
>> I also had two test crashes:
>>
>> org.apache.harmony.luni.tests.java.net.SocketTest
>> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>>
>> which were both fixed by Ilya's patch in HARMONY-6132.
>>
>> I think it would be good to apply the patch from HARMONY-6132 before M10.
>> Does anyone else support/oppose this?
>>
>
> The fix looks reasonable and the issue is important enough to fix it in M10.
> --
> Alexey
>
>
>> Regards,
>> Mark.
>>
>>
>>
>>
>
>
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
Re: [M10] Testing
Posted by Alexey Varlamov <al...@gmail.com>.
2009/5/28, Mark Hindess <ma...@googlemail.com>:
[skipped]
> I also had two test crashes:
>
> org.apache.harmony.luni.tests.java.net.SocketTest
> org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
>
> which were both fixed by Ilya's patch in HARMONY-6132.
>
> I think it would be good to apply the patch from HARMONY-6132 before M10.
> Does anyone else support/oppose this?
The fix looks reasonable and the issue is important enough to fix it in M10.
--
Alexey
>
> Regards,
> Mark.
>
>
>
Re: [M10] Testing
Posted by Mark Hindess <ma...@googlemail.com>.
In message <20...@d12av04.megacenter.de.ibm.com>,
Mark Hindess writes:
>
> In message <4A...@googlemail.com>, Oliver Deakin writes:
> >
> > Hi all,
> >
> > I've just run the classlib tests on Windows x86 on the latest code in
> > the federated build and I don't see anything catastrophic there. There's
> > 1 failure in the portlib tests, in hymmap, which I think is a test case
> > off-by-one error. I've attached a patch [1] which fixes the test. Apart
> > from that we look clean.
> >
> > Does anyone have any other test reports? Any other platforms? I'm
> > assuming we're taking r778555 as our testing base for M10.
>
> I think we said we were going to test building from source this time
> around so I'm testing that:
>
>
> ant bundle-src
> tar xzf target/apache-harmony-src-r778555-snapshot.tar.gz
> cd apache-harmony-src-r778555
> ant fetch-depends snapshot
> ant -Dtest.jre.home=$PWD/target/hdk/jdk/jre -f working_classlib/build.xml test
>
> works on Linux. We need to check that this builds a sane binary
> snapshot with the correct version information in the various jars, etc.
>
> I'll report back once my test run is complete.
Well, I had some test failures caused by timeouts in BasicSwingTestCase.
For example:
javax.swing.BoxLayoutTest.testLayoutContainerVertical
Failure Test interrupted due timeout 10.094
junit.framework.AssertionFailedError: Test interrupted due timeout
at javax.swing.BasicSwingTestCase.runBare (BasicSwingTestCase.java:147)
but I've seen these before on my laptop so they aren't blockers for M10.
I also had two test crashes:
org.apache.harmony.luni.tests.java.net.SocketTest
org.apache.harmony.xnet.provider.jsse.SSLSocketImplTest
which were both fixed by Ilya's patch in HARMONY-6132.
I think it would be good to apply the patch from HARMONY-6132 before M10.
Does anyone else support/oppose this?
Regards,
Mark.
Re: [M10] Testing
Posted by Oliver Deakin <ol...@googlemail.com>.
Mark Hindess wrote:
> Oliver, regarding your patch below, since the + 1 is for the trailing \0
> (rather than the separator) perhaps we should move it to the end of the
> sum rather than have it in the middle?
>
Yep, will do when I apply it.
Regards,
Oliver
Re: [M10] Testing
Posted by Mark Hindess <ma...@googlemail.com>.
In message <4A...@googlemail.com>, Oliver Deakin writes:
>
> Hi all,
>
> I've just run the classlib tests on Windows x86 on the latest code in
> the federated build and I don't see anything catastrophic there. There's
> 1 failure in the portlib tests, in hymmap, which I think is a test case
> off-by-one error. I've attached a patch [1] which fixes the test. Apart
> from that we look clean.
>
> Does anyone have any other test reports? Any other platforms? I'm
> assuming we're taking r778555 as our testing base for M10.
I think we said we were going to test building from source this time
around so I'm testing that:
ant bundle-src
tar xzf target/apache-harmony-src-r778555-snapshot.tar.gz
cd apache-harmony-src-r778555
ant fetch-depends snapshot
ant -Dtest.jre.home=$PWD/target/hdk/jdk/jre -f working_classlib/build.xml test
works on Linux. We need to check that this builds a sane binary
snapshot with the correct version information in the various jars, etc.
I'll report back once my test run is complete.
Oliver, regarding your patch below, since the + 1 is for the trailing \0
(rather than the separator) perhaps we should move it to the end of the
sum rather than have it in the middle?
Regards,
Mark.
> Regards,
> Oliver
>
> [1]
> Adding 2 to the path size for emptyFile and testFile is incorrect, as
> pathLen already counts the rightmost DIR_SEPARATOR.
>
> Index: modules/portlib/src/test/native/hymmap/shared/hymmap.c
> ===================================================================
> --- modules/portlib/src/test/native/hymmap/shared/hymmap.c (revision
> 779066)
> +++ modules/portlib/src/test/native/hymmap/shared/hymmap.c (working copy)
> @@ -126,8 +126,8 @@
> execName = strrchr(execPath, DIR_SEPARATOR) + 1; /* Find the
> rightmost slash character */
> pathLen = strlen(execPath) - strlen(execName);
>
> - emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
> pathLen + strlen("shared") + 2 + strlen("emptyFile")); /* +2 for the
> extra slash and null terminator */
> - testFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
> pathLen + strlen("shared") + 2 + strlen("testFile"));
> + emptyFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
> pathLen + strlen("shared") + 1 + strlen("emptyFile")); /* +1 for the
> null terminator */
> + testFile = hyportLibrary->mem_allocate_memory(hyportLibrary,
> pathLen + strlen("shared") + 1 + strlen("testFile"));
>
> strncpy(emptyFile, execPath, pathLen);
> strcat(emptyFile, "shared");