* Has rgw recently switched to deque usuage??
@ 2016-02-20 14:41 Willem Jan Withagen
2016-02-20 21:33 ` Matt Benjamin
0 siblings, 1 reply; 9+ messages in thread
From: Willem Jan Withagen @ 2016-02-20 14:41 UTC (permalink / raw)
To: Ceph Development
Hi,
With the last rebase I'm getting compilation errors on rgw stuff.
So I was wondering if deque usuage got included recently, or that I'm
running into a case where Clang doesn't like the way some code is used.
And GCC just compiles fine...
Any answer would be appriciated
--WjW
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-20 14:41 Has rgw recently switched to deque usuage?? Willem Jan Withagen
@ 2016-02-20 21:33 ` Matt Benjamin
2016-02-20 22:15 ` Willem Jan Withagen
0 siblings, 1 reply; 9+ messages in thread
From: Matt Benjamin @ 2016-02-20 21:33 UTC (permalink / raw)
To: Willem Jan Withagen; +Cc: Ceph Development
Hi Willem,
A number of new features have merged, sure, and they make use of new primitives. I haven't compiled with clang recently, mainly due to issues linking with gcc-compiled leveldb and rocksdb on Linux platforms. For now, maybe you can send more specific info.
(I'll be able to try-build on FreeBSD at some point, probably, but not in the next day or so due to travel.)
Matt
----- Original Message -----
> From: "Willem Jan Withagen" <wjw@digiware.nl>
> To: "Ceph Development" <ceph-devel@vger.kernel.org>
> Sent: Saturday, February 20, 2016 9:41:45 AM
> Subject: Has rgw recently switched to deque usuage??
>
> Hi,
>
> With the last rebase I'm getting compilation errors on rgw stuff.
> So I was wondering if deque usuage got included recently, or that I'm
> running into a case where Clang doesn't like the way some code is used.
> And GCC just compiles fine...
>
> Any answer would be appriciated
>
> --WjW
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
--
Matt Benjamin
Red Hat, Inc.
315 West Huron Street, Suite 140A
Ann Arbor, Michigan 48103
http://www.redhat.com/en/technologies/storage
tel. 734-707-0660
fax. 734-769-8938
cel. 734-216-5309
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-20 21:33 ` Matt Benjamin
@ 2016-02-20 22:15 ` Willem Jan Withagen
2016-02-21 0:13 ` Matt Benjamin
0 siblings, 1 reply; 9+ messages in thread
From: Willem Jan Withagen @ 2016-02-20 22:15 UTC (permalink / raw)
To: Matt Benjamin; +Cc: Ceph Development
On 20-2-2016 22:33, Matt Benjamin wrote:
> Hi Willem,
>
> A number of new features have merged, sure, and they make use of new
> primitives. I haven't compiled with clang recently, mainly due to
> issues linking with gcc-compiled leveldb and rocksdb on Linux
> platforms. For now, maybe you can send more specific info.
>
> (I'll be able to try-build on FreeBSD at some point, probably, but
> not in the next day or so due to travel.)
Hi Matt,
It turns out to be all about the order of includes and that Clang not
always sees eye to eye with GCC.
Right now I'm sort of reordering a lot includes rados/rgw headers in the
rgw/*cc files.
Probably too much reordering but first I want it to get it to compile.
:) After that lets see what was really needed.
Save travels,
--WjW
> Matt
>
> ----- Original Message -----
>> From: "Willem Jan Withagen" <wjw@digiware.nl> To: "Ceph
>> Development" <ceph-devel@vger.kernel.org> Sent: Saturday, February
>> 20, 2016 9:41:45 AM Subject: Has rgw recently switched to deque
>> usuage??
>>
>> Hi,
>>
>> With the last rebase I'm getting compilation errors on rgw stuff.
>> So I was wondering if deque usuage got included recently, or that
>> I'm running into a case where Clang doesn't like the way some code
>> is used. And GCC just compiles fine...
>>
>> Any answer would be appriciated
>>
>> --WjW -- To unsubscribe from this list: send the line "unsubscribe
>> ceph-devel" in the body of a message to majordomo@vger.kernel.org
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-20 22:15 ` Willem Jan Withagen
@ 2016-02-21 0:13 ` Matt Benjamin
2016-02-21 0:49 ` Willem Jan Withagen
2016-02-21 10:46 ` Willem Jan Withagen
0 siblings, 2 replies; 9+ messages in thread
From: Matt Benjamin @ 2016-02-21 0:13 UTC (permalink / raw)
To: Willem Jan Withagen; +Cc: Ceph Development
Hi Willem,
That's good news, but massive header re-ordering sounds suspicious. Likely, there are just missing unix headers.
Matt
----- Original Message -----
> From: "Willem Jan Withagen" <wjw@digiware.nl>
> To: "Matt Benjamin" <mbenjamin@redhat.com>
> Cc: "Ceph Development" <ceph-devel@vger.kernel.org>
> Sent: Saturday, February 20, 2016 5:15:44 PM
> Subject: Re: Has rgw recently switched to deque usuage??
>
> On 20-2-2016 22:33, Matt Benjamin wrote:
> > Hi Willem,
> >
> > A number of new features have merged, sure, and they make use of new
> > primitives. I haven't compiled with clang recently, mainly due to
> > issues linking with gcc-compiled leveldb and rocksdb on Linux
> > platforms. For now, maybe you can send more specific info.
> >
> > (I'll be able to try-build on FreeBSD at some point, probably, but
> > not in the next day or so due to travel.)
>
> Hi Matt,
>
> It turns out to be all about the order of includes and that Clang not
> always sees eye to eye with GCC.
> Right now I'm sort of reordering a lot includes rados/rgw headers in the
> rgw/*cc files.
> Probably too much reordering but first I want it to get it to compile.
> :) After that lets see what was really needed.
>
> Save travels,
> --WjW
>
> > Matt
> >
> > ----- Original Message -----
> >> From: "Willem Jan Withagen" <wjw@digiware.nl> To: "Ceph
> >> Development" <ceph-devel@vger.kernel.org> Sent: Saturday, February
> >> 20, 2016 9:41:45 AM Subject: Has rgw recently switched to deque
> >> usuage??
> >>
> >> Hi,
> >>
> >> With the last rebase I'm getting compilation errors on rgw stuff.
> >> So I was wondering if deque usuage got included recently, or that
> >> I'm running into a case where Clang doesn't like the way some code
> >> is used. And GCC just compiles fine...
> >>
> >> Any answer would be appriciated
> >>
> >> --WjW -- To unsubscribe from this list: send the line "unsubscribe
> >> ceph-devel" in the body of a message to majordomo@vger.kernel.org
> >> More majordomo info at http://vger.kernel.org/majordomo-info.html
> >>
> >
>
>
--
--
Matt Benjamin
Red Hat, Inc.
315 West Huron Street, Suite 140A
Ann Arbor, Michigan 48103
http://www.redhat.com/en/technologies/storage
tel. 734-707-0660
fax. 734-769-8938
cel. 734-216-5309
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-21 0:13 ` Matt Benjamin
@ 2016-02-21 0:49 ` Willem Jan Withagen
2016-02-21 10:46 ` Willem Jan Withagen
1 sibling, 0 replies; 9+ messages in thread
From: Willem Jan Withagen @ 2016-02-21 0:49 UTC (permalink / raw)
To: Matt Benjamin; +Cc: Ceph Development
On 21-2-2016 01:13, Matt Benjamin wrote:
> Hi Willem,
>
> That's good news, but massive header re-ordering sounds suspicious. Likely, there are just missing unix headers.
Well perhaps massive is the wrong word. And Clang is a lot more picky
than GCC....
Major problem is that compiling rgw modules complains with just about
all modules:
In file included from test/rgw/test_rgw_period_history.cc:14:
./rgw/rgw_period_history.h:24:16: error: use of undeclared identifier
'RGWPeriod'
std::deque<RGWPeriod> periods;
^
Which is relieved by including:
#include "rgw_rados.h"
#include "rgw_period_puller.h"
#include "rgw_period_history.h"
#include "rgw_metadata.h"
In all complaining modules.
And needed to fix including ordering, where std::queue was called on
things not yet fully defined.
same for the modules in src/test....
--WjW
>
> Matt
>
> ----- Original Message -----
>> From: "Willem Jan Withagen" <wjw@digiware.nl>
>> To: "Matt Benjamin" <mbenjamin@redhat.com>
>> Cc: "Ceph Development" <ceph-devel@vger.kernel.org>
>> Sent: Saturday, February 20, 2016 5:15:44 PM
>> Subject: Re: Has rgw recently switched to deque usuage??
>>
>> On 20-2-2016 22:33, Matt Benjamin wrote:
>>> Hi Willem,
>>>
>>> A number of new features have merged, sure, and they make use of new
>>> primitives. I haven't compiled with clang recently, mainly due to
>>> issues linking with gcc-compiled leveldb and rocksdb on Linux
>>> platforms. For now, maybe you can send more specific info.
>>>
>>> (I'll be able to try-build on FreeBSD at some point, probably, but
>>> not in the next day or so due to travel.)
>>
>> Hi Matt,
>>
>> It turns out to be all about the order of includes and that Clang not
>> always sees eye to eye with GCC.
>> Right now I'm sort of reordering a lot includes rados/rgw headers in the
>> rgw/*cc files.
>> Probably too much reordering but first I want it to get it to compile.
>> :) After that lets see what was really needed.
>>
>> Save travels,
>> --WjW
>>
>>> Matt
>>>
>>> ----- Original Message -----
>>>> From: "Willem Jan Withagen" <wjw@digiware.nl> To: "Ceph
>>>> Development" <ceph-devel@vger.kernel.org> Sent: Saturday, February
>>>> 20, 2016 9:41:45 AM Subject: Has rgw recently switched to deque
>>>> usuage??
>>>>
>>>> Hi,
>>>>
>>>> With the last rebase I'm getting compilation errors on rgw stuff.
>>>> So I was wondering if deque usuage got included recently, or that
>>>> I'm running into a case where Clang doesn't like the way some code
>>>> is used. And GCC just compiles fine...
>>>>
>>>> Any answer would be appriciated
>>>>
>>>> --WjW -- To unsubscribe from this list: send the line "unsubscribe
>>>> ceph-devel" in the body of a message to majordomo@vger.kernel.org
>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>>>
>>>
>>
>>
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-21 0:13 ` Matt Benjamin
2016-02-21 0:49 ` Willem Jan Withagen
@ 2016-02-21 10:46 ` Willem Jan Withagen
2016-02-21 11:00 ` Matt Benjamin
1 sibling, 1 reply; 9+ messages in thread
From: Willem Jan Withagen @ 2016-02-21 10:46 UTC (permalink / raw)
To: Matt Benjamin; +Cc: Ceph Development
On 21-2-2016 01:13, Matt Benjamin wrote:
> Hi Willem,
>
> That's good news, but massive header re-ordering sounds suspicious. Likely, there are just missing unix headers.
Overnight the build completed.
Now running gmake check.
If that gets to my normal level of PASSED of around 110 of 133 than I
commit the rgw changes and you can find that in my pull to comment about
and we take it from there.
--WjW
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-21 10:46 ` Willem Jan Withagen
@ 2016-02-21 11:00 ` Matt Benjamin
2016-02-21 13:18 ` Willem Jan Withagen
2016-02-22 15:12 ` RGW include modifications to compile with Clang Willem Jan Withagen
0 siblings, 2 replies; 9+ messages in thread
From: Matt Benjamin @ 2016-02-21 11:00 UTC (permalink / raw)
To: Willem Jan Withagen; +Cc: Ceph Development
sure, send github pull request tagged "rgw" when ready
Matt
----- Original Message -----
> From: "Willem Jan Withagen" <wjw@digiware.nl>
> To: "Matt Benjamin" <mbenjamin@redhat.com>
> Cc: "Ceph Development" <ceph-devel@vger.kernel.org>
> Sent: Sunday, February 21, 2016 5:46:38 AM
> Subject: Re: Has rgw recently switched to deque usuage??
>
> On 21-2-2016 01:13, Matt Benjamin wrote:
> > Hi Willem,
> >
> > That's good news, but massive header re-ordering sounds suspicious.
> > Likely, there are just missing unix headers.
>
> Overnight the build completed.
> Now running gmake check.
> If that gets to my normal level of PASSED of around 110 of 133 than I
> commit the rgw changes and you can find that in my pull to comment about
> and we take it from there.
>
> --WjW
>
>
--
--
Matt Benjamin
Red Hat, Inc.
315 West Huron Street, Suite 140A
Ann Arbor, Michigan 48103
http://www.redhat.com/en/technologies/storage
tel. 734-707-0660
fax. 734-769-8938
cel. 734-216-5309
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Has rgw recently switched to deque usuage??
2016-02-21 11:00 ` Matt Benjamin
@ 2016-02-21 13:18 ` Willem Jan Withagen
2016-02-22 15:12 ` RGW include modifications to compile with Clang Willem Jan Withagen
1 sibling, 0 replies; 9+ messages in thread
From: Willem Jan Withagen @ 2016-02-21 13:18 UTC (permalink / raw)
To: Matt Benjamin; +Cc: Ceph Development
On 21-2-2016 12:00, Matt Benjamin wrote:
> sure, send github pull request tagged "rgw" when ready
Second try in
https://github.com/ceph/ceph/pull/7730
First one was on top of all my other FreeBSD changes, so I
tried making a new one passed on my master checkout.
But yoy still get all the other commits.
So you need to look at commit 8c4530f
--WjW
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: RGW include modifications to compile with Clang
2016-02-21 11:00 ` Matt Benjamin
2016-02-21 13:18 ` Willem Jan Withagen
@ 2016-02-22 15:12 ` Willem Jan Withagen
1 sibling, 0 replies; 9+ messages in thread
From: Willem Jan Withagen @ 2016-02-22 15:12 UTC (permalink / raw)
To: Matt Benjamin; +Cc: Ceph Development
On 21-2-2016 12:00, Matt Benjamin wrote:
> sure, send github pull request tagged "rgw" when ready
>
Hi,
It needed some hair pulling....
But the last for commits of #7730 generate a compilable version, for
both FreeBSD/Clang and Linux/GCC, for as far as I can see.
--WjW
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2016-02-22 15:12 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-02-20 14:41 Has rgw recently switched to deque usuage?? Willem Jan Withagen
2016-02-20 21:33 ` Matt Benjamin
2016-02-20 22:15 ` Willem Jan Withagen
2016-02-21 0:13 ` Matt Benjamin
2016-02-21 0:49 ` Willem Jan Withagen
2016-02-21 10:46 ` Willem Jan Withagen
2016-02-21 11:00 ` Matt Benjamin
2016-02-21 13:18 ` Willem Jan Withagen
2016-02-22 15:12 ` RGW include modifications to compile with Clang Willem Jan Withagen
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.