threads / discuss / 40410

git-p4: t9819 failing

Subject: git-p4: t9819 failing

## tl;dr

7 messages between Sep 23, 2015 and Sep 25, 2015.

replies: 6people: 2as markdown or json

Luke Diamand· Sep 23, 2015, 09:27 UTC · lore

Lars,

I just noticed that your change "git-p4: honor core.ignorecase when using P4 client specs" seems to break t9819.

I suspect that the problem is just that t9819 needs to be updated to reflect your change - do you have any thoughts on that?

Thanks! Luke

Lars Schneider· Sep 23, 2015, 10:27 UTC · re: Luke Diamand · lore

Re: git-p4: t9819 failing

On 23 Sep 2015, at 11:27, Luke Diamand <luke@diamand.org> wrote:

Show 10 quoted lines

> Lars,
> 
> I just noticed that your change "git-p4: honor core.ignorecase when
> using P4 client specs" seems to break t9819.
> 
> I suspect that the problem is just that t9819 needs to be updated to
> reflect your change - do you have any thoughts on that?
> 
> Thanks!
> Luke

What OS and what commit do you use for testing? Here is what I got on my machine:

Ubuntu, next (c07a1e8) -> OK Ubuntu, maint (ee6ad5f) -> OK

OS X, next (c07a1e8) -> broken OS X, maint (ee6ad5f) -> broken

Thanks, Lars

Luke Diamand· Sep 23, 2015, 11:11 UTC · re: Lars Schneider · lore

Re: git-p4: t9819 failing

On 23 September 2015 at 11:27, Lars Schneider <larsxschneider@gmail.com> wrote:

Show 21 quoted lines

>
> On 23 Sep 2015, at 11:27, Luke Diamand <luke@diamand.org> wrote:
>
>> Lars,
>>
>> I just noticed that your change "git-p4: honor core.ignorecase when
>> using P4 client specs" seems to break t9819.
>>
>> I suspect that the problem is just that t9819 needs to be updated to
>> reflect your change - do you have any thoughts on that?
>>
>> Thanks!
>> Luke
>
> What OS and what commit do you use for testing? Here is what I got on my machine:
>
> Ubuntu, next (c07a1e8) -> OK
> Ubuntu, maint (ee6ad5f) -> OK
>
> OS X, next (c07a1e8) -> broken
> OS X, maint (ee6ad5f) -> broken

The next branch is good: c07a1e8 Merge branch 'bb/remote-get-url' into next

The pu branch isn't: f44e3df Merge branch 'jk/notes-dwim-doc' into pu

This is on Debian.

Actually I think the problem may be in calcDiskFree().

Here's the last bit of the crash dump from git-p4 I get:

  File "/home/ldiamand/git/git/git-p4", line 2580, in streamP4FilesCbSelf
    self.streamP4FilesCb(entry)
  File "/home/ldiamand/git/git/git-p4", line 2497, in streamP4FilesCb
    required_bytes = int((4 * int(self.stream_file["fileSize"])) -
calcDiskFree(self.cloneDestination))
  File "/home/ldiamand/git/git/git-p4", line 116, in calcDiskFree
    st = os.statvfs(dirname)
OSError: [Errno 2] No such file or directory: 'lc'

Luke

Lars Schneider· Sep 23, 2015, 12:28 UTC · re: Luke Diamand · lore

Re: git-p4: t9819 failing

On 23 Sep 2015, at 13:11, Luke Diamand <luke@diamand.org> wrote:

Show 45 quoted lines

> On 23 September 2015 at 11:27, Lars Schneider <larsxschneider@gmail.com> wrote:
>> 
>> On 23 Sep 2015, at 11:27, Luke Diamand <luke@diamand.org> wrote:
>> 
>>> Lars,
>>> 
>>> I just noticed that your change "git-p4: honor core.ignorecase when
>>> using P4 client specs" seems to break t9819.
>>> 
>>> I suspect that the problem is just that t9819 needs to be updated to
>>> reflect your change - do you have any thoughts on that?
>>> 
>>> Thanks!
>>> Luke
>> 
>> What OS and what commit do you use for testing? Here is what I got on my machine:
>> 
>> Ubuntu, next (c07a1e8) -> OK
>> Ubuntu, maint (ee6ad5f) -> OK
>> 
>> OS X, next (c07a1e8) -> broken
>> OS X, maint (ee6ad5f) -> broken
> 
> The next branch is good:
> c07a1e8 Merge branch 'bb/remote-get-url' into next
> 
> The pu branch isn't:
> f44e3df Merge branch 'jk/notes-dwim-doc' into pu
> 
> This is on Debian.
> 
> Actually I think the problem may be in calcDiskFree().
> 
> Here's the last bit of the crash dump from git-p4 I get:
> 
>  File "/home/ldiamand/git/git/git-p4", line 2580, in streamP4FilesCbSelf
>    self.streamP4FilesCb(entry)
>  File "/home/ldiamand/git/git/git-p4", line 2497, in streamP4FilesCb
>    required_bytes = int((4 * int(self.stream_file["fileSize"])) -
> calcDiskFree(self.cloneDestination))
>  File "/home/ldiamand/git/git/git-p4", line 116, in calcDiskFree
>    st = os.statvfs(dirname)
> OSError: [Errno 2] No such file or directory: 'lc'
> 
> Luke

Confirmed. What do you think about this fix?

Thank you, Lars

---
 git-p4.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/git-p4.py b/git-p4.py
index 1d1bb87..66c0a4e 100755
--- a/git-p4.py
+++ b/git-p4.py
@@ -3478,6 +3478,7 @@ class P4Clone(P4Sync):
 
         print "Importing from %s into %s" % (', '.join(depotPaths), self.cloneDestination)
 
+        self.cloneDestination = os.path.abspath(self.cloneDestination)
         if not os.path.exists(self.cloneDestination):
             os.makedirs(self.cloneDestination)
         chdir(self.cloneDestination)

Luke Diamand· Sep 24, 2015, 07:45 UTC · re: Lars Schneider · lore

Re: git-p4: t9819 failing

On 23 September 2015 at 13:28, Lars Schneider <larsxschneider@gmail.com> wrote:

Show 14 quoted lines

>
>> Here's the last bit of the crash dump from git-p4 I get:
>>
>>  File "/home/ldiamand/git/git/git-p4", line 2580, in streamP4FilesCbSelf
>>    self.streamP4FilesCb(entry)
>>  File "/home/ldiamand/git/git/git-p4", line 2497, in streamP4FilesCb
>>    required_bytes = int((4 * int(self.stream_file["fileSize"])) -
>> calcDiskFree(self.cloneDestination))
>>  File "/home/ldiamand/git/git/git-p4", line 116, in calcDiskFree
>>    st = os.statvfs(dirname)
>> OSError: [Errno 2] No such file or directory: 'lc'
>>
>> Luke
> Confirmed. What do you think about this fix?

Works for me!

Show 21 quoted lines

>
> Thank you,
> Lars
>
> ---
>  git-p4.py | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/git-p4.py b/git-p4.py
> index 1d1bb87..66c0a4e 100755
> --- a/git-p4.py
> +++ b/git-p4.py
> @@ -3478,6 +3478,7 @@ class P4Clone(P4Sync):
>
>          print "Importing from %s into %s" % (', '.join(depotPaths), self.cloneDestination)
>
> +        self.cloneDestination = os.path.abspath(self.cloneDestination)
>          if not os.path.exists(self.cloneDestination):
>              os.makedirs(self.cloneDestination)
>          chdir(self.cloneDestination)
> --

Luke Diamand· Sep 24, 2015, 20:29 UTC · re: Luke Diamand · lore

Re: git-p4: t9819 failing

OK, slight correction there - it now doesn't crash getting the disk usage, but I think it still needs to be updated following the other changes to case-handling.

Luke

On 24 September 2015 at 08:45, Luke Diamand <luke@diamand.org> wrote:

Show 41 quoted lines

> On 23 September 2015 at 13:28, Lars Schneider <larsxschneider@gmail.com> wrote:
>>
>>> Here's the last bit of the crash dump from git-p4 I get:
>>>
>>>  File "/home/ldiamand/git/git/git-p4", line 2580, in streamP4FilesCbSelf
>>>    self.streamP4FilesCb(entry)
>>>  File "/home/ldiamand/git/git/git-p4", line 2497, in streamP4FilesCb
>>>    required_bytes = int((4 * int(self.stream_file["fileSize"])) -
>>> calcDiskFree(self.cloneDestination))
>>>  File "/home/ldiamand/git/git/git-p4", line 116, in calcDiskFree
>>>    st = os.statvfs(dirname)
>>> OSError: [Errno 2] No such file or directory: 'lc'
>>>
>>> Luke
>> Confirmed. What do you think about this fix?
>
> Works for me!
>
>
>
>>
>> Thank you,
>> Lars
>>
>> ---
>>  git-p4.py | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/git-p4.py b/git-p4.py
>> index 1d1bb87..66c0a4e 100755
>> --- a/git-p4.py
>> +++ b/git-p4.py
>> @@ -3478,6 +3478,7 @@ class P4Clone(P4Sync):
>>
>>          print "Importing from %s into %s" % (', '.join(depotPaths), self.cloneDestination)
>>
>> +        self.cloneDestination = os.path.abspath(self.cloneDestination)
>>          if not os.path.exists(self.cloneDestination):
>>              os.makedirs(self.cloneDestination)
>>          chdir(self.cloneDestination)
>> --

Lars Schneider· Sep 25, 2015, 09:29 UTC · re: Luke Diamand · lore

Re: git-p4: t9819 failing

Good catch. The case-handling test is actually fine. There was a bug in my implementation!

If you do this: git p4 clone //depot/something/...

Then git p4 generates a directory “something” and clones into that (similar to Git). Since I set “cloneDirectory” to the current working directory that logic never kicked in. Therefore the depot was cloned into the current working directory instead of a new directory “something” and the test broke.

Thanks, Lars

On 24 Sep 2015, at 22:29, Luke Diamand <luke@diamand.org> wrote:

Show 48 quoted lines

> OK, slight correction there - it now doesn't crash getting the disk
> usage, but I think it still needs to be updated following the other
> changes to case-handling.
> 
> Luke
> 
> On 24 September 2015 at 08:45, Luke Diamand <luke@diamand.org> wrote:
>> On 23 September 2015 at 13:28, Lars Schneider <larsxschneider@gmail.com> wrote:
>>> 
>>>> Here's the last bit of the crash dump from git-p4 I get:
>>>> 
>>>> File "/home/ldiamand/git/git/git-p4", line 2580, in streamP4FilesCbSelf
>>>>   self.streamP4FilesCb(entry)
>>>> File "/home/ldiamand/git/git/git-p4", line 2497, in streamP4FilesCb
>>>>   required_bytes = int((4 * int(self.stream_file["fileSize"])) -
>>>> calcDiskFree(self.cloneDestination))
>>>> File "/home/ldiamand/git/git/git-p4", line 116, in calcDiskFree
>>>>   st = os.statvfs(dirname)
>>>> OSError: [Errno 2] No such file or directory: 'lc'
>>>> 
>>>> Luke
>>> Confirmed. What do you think about this fix?
>> 
>> Works for me!
>> 
>> 
>> 
>>> 
>>> Thank you,
>>> Lars
>>> 
>>> ---
>>> git-p4.py | 1 +
>>> 1 file changed, 1 insertion(+)
>>> 
>>> diff --git a/git-p4.py b/git-p4.py
>>> index 1d1bb87..66c0a4e 100755
>>> --- a/git-p4.py
>>> +++ b/git-p4.py
>>> @@ -3478,6 +3478,7 @@ class P4Clone(P4Sync):
>>> 
>>>         print "Importing from %s into %s" % (', '.join(depotPaths), self.cloneDestination)
>>> 
>>> +        self.cloneDestination = os.path.abspath(self.cloneDestination)
>>>         if not os.path.exists(self.cloneDestination):
>>>             os.makedirs(self.cloneDestination)
>>>         chdir(self.cloneDestination)
>>> --