From: Willem Jan Withagen <wjw@digiware.nl>
To: Sage Weil <sage@newdream.net>, Mykola Golub <mgolub@mirantis.com>
Cc: "Yan, Zheng" <ukernel@gmail.com>, Haomai Wang <haomai@xsky.com>,
Ceph Development <ceph-devel@vger.kernel.org>
Subject: Re: Compiling for FreeBSD, Bluestore requires AIO
Date: Fri, 15 Jan 2016 12:21:08 +0100 [thread overview]
Message-ID: <5698D624.6000301@digiware.nl> (raw)
In-Reply-To: <5698CF58.7020206@digiware.nl>
On 15-1-2016 11:52, Willem Jan Withagen wrote:
> On 30-11-2015 14:21, Sage Weil wrote:
>> The problem with all of the porting code in general is that it is doomed
>> to break later on if we don't have (at least) ongoing build tests. In
>> order for a FreeBSD or OSX port to continue working we need VMs that run
>> either gitbuilder or a jenkins job or similar so that we can tell when it
>> breaks.
>>
>> If someone is willing to run a VM somewhere to do this we can pretty
>> easily stick it on the gitbuilder page at
>>
>> http://ceph.com/gitbuilder.cgi
>
> Well this is real nice case of such an incident.
>
> I'm verifying my changes to the C/C++ code, after having forwarded to HEAD.
> And now I get bluestore complaining that it requires AIO.
> Something I've thusfar excluded in the build.
>
> Does Bluestore really require AIO??
> FreeBSD does have AIO, but to take simple steps one at the time, I
> excluded it
> with configure. Mainly because it bombs at compile time. Probably due to
> different include files, or function nameing.....
>
> So, in order of preference:
> - Can I disable AIO for Bluestore
> - Can I disable bluestore being build?
Spoke a bit too soon.
FreeBSD has AIO, loadable as kernelmodule. with the following systemcalls.
aio_cancel(2), aio_error(2), aio_read(2), aio_return(2), aio_suspend(2),
aio_waitcomplete(2), aio_write(2), lio_listio(2)
But libaio is based on:
/* Actual syscalls */
extern int io_setup(int maxevents, io_context_t *ctxp);
extern int io_destroy(io_context_t ctx);
extern int io_submit(io_context_t ctx, long nr, struct iocb *ios[]);
extern int io_cancel(io_context_t ctx, struct iocb *iocb, struct
io_event *evt);
extern int io_getevents(io_context_t ctx_id, long min_nr, long nr,
struct io_event *events, struct timespec *timeout);
So glueing that together is not just fixing some include files, and or
fetching
some shim-library that is lurking out in /usr/ports...
And for the time being I would like to skip over the libaio hurdle.
Is that possible?
--WjW
next prev parent reply other threads:[~2016-01-17 16:30 UTC|newest]
Thread overview: 61+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-11-29 17:44 Compiling for FreeBSD Willem Jan Withagen
2015-11-29 18:08 ` Haomai Wang
2015-11-29 18:57 ` Willem Jan Withagen
2015-11-30 3:46 ` Yan, Zheng
2015-11-30 6:58 ` Mykola Golub
2015-11-30 11:53 ` Willem Jan Withagen
2015-11-30 14:13 ` Mykola Golub
2015-11-30 14:40 ` Willem Jan Withagen
2015-11-30 16:04 ` Willem Jan Withagen
2015-11-30 16:20 ` Gregory Farnum
2015-12-01 9:42 ` Willem Jan Withagen
2015-12-01 12:22 ` Mykola Golub
2015-12-01 12:44 ` Willem Jan Withagen
2015-11-30 14:56 ` Willem Jan Withagen
2015-11-30 13:21 ` Sage Weil
2015-11-30 13:54 ` Willem Jan Withagen
2015-12-01 11:08 ` Willem Jan Withagen
2015-12-01 13:30 ` Sage Weil
2015-12-01 13:42 ` Willem Jan Withagen
2015-12-01 14:35 ` Sage Weil
2015-12-01 16:24 ` Willem Jan Withagen
2015-12-01 17:22 ` Alan Somers
2015-12-01 18:08 ` Willem Jan Withagen
2015-12-01 18:21 ` Alan Somers
2015-12-01 18:31 ` Willem Jan Withagen
2015-12-01 18:36 ` Sage Weil
2015-12-01 18:43 ` Willem Jan Withagen
2015-12-02 14:13 ` Yan, Zheng
2015-12-02 20:52 ` Willem Jan Withagen
2015-12-03 0:27 ` Yan, Zheng
2015-12-04 18:30 ` Willem Jan Withagen
2015-12-04 18:44 ` Gregory Farnum
2015-12-04 20:11 ` Willem Jan Withagen
2015-12-08 9:59 ` Willem Jan Withagen
2015-12-08 10:04 ` Willem Jan Withagen
2015-12-08 12:36 ` Mykola Golub
2015-12-08 15:59 ` Willem Jan Withagen
2015-12-01 18:51 ` Willem Jan Withagen
2015-12-02 21:10 ` Willem Jan Withagen
2015-12-02 22:47 ` Compiling for FreeBSD, missing rbd Willem Jan Withagen
2015-12-03 12:34 ` Mykola Golub
2015-12-03 13:27 ` Willem Jan Withagen
2015-12-03 9:50 ` Compiling for FreeBSD, runtimes for seperate tests Willem Jan Withagen
2015-12-03 14:12 ` Willem Jan Withagen
2015-12-03 21:06 ` Gregory Farnum
2015-12-05 12:56 ` Compiling for FreeBSD, Clang refuses to compile a test Willem Jan Withagen
2015-12-05 13:02 ` Xinze Chi (信泽)
2015-12-07 21:44 ` Willem Jan Withagen
2015-12-07 22:19 ` Michal Jarzabek
2015-12-08 0:29 ` Willem Jan Withagen
2015-12-08 8:48 ` Willem Jan Withagen
2016-01-16 12:56 ` Compiling for FreeBSD Willem Jan Withagen
2015-12-10 15:03 ` Compiling for FreeBSD, trouble in buffer.c Willem Jan Withagen
2015-12-11 9:56 ` Willem Jan Withagen
2016-01-15 10:52 ` Compiling for FreeBSD, Bluestore requires AIO Willem Jan Withagen
2016-01-15 11:21 ` Willem Jan Withagen [this message]
2016-01-15 17:30 ` Sage Weil
2016-01-15 18:34 ` Willem Jan Withagen
2016-01-18 9:54 ` Mykola Golub
2016-01-18 10:05 ` Willem Jan Withagen
2016-05-28 0:15 ` Compiling for FreeBSD Willem Jan Withagen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5698D624.6000301@digiware.nl \
--to=wjw@digiware.nl \
--cc=ceph-devel@vger.kernel.org \
--cc=haomai@xsky.com \
--cc=mgolub@mirantis.com \
--cc=sage@newdream.net \
--cc=ukernel@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox