> On Oct. 18, 2012, 5:09 p.m., Ben Mahler wrote:
> > src/linux/cgroups.cpp, line 194
> > <https://reviews.apache.org/r/7620/diff/1/?file=177249#file177249line194>
> >
> >     Any reason for the 'cgroups::' prefix on the read/writeControl calls 
> > here?

Bad copy-pasta, thanks.


> On Oct. 18, 2012, 5:09 p.m., Ben Mahler wrote:
> > src/linux/cgroups.cpp, line 256
> > <https://reviews.apache.org/r/7620/diff/1/?file=177249#file177249line256>
> >
> >     1. The call to path::join("/", cgroup) is quite strange, it will 
> > essentially just prepend a "/" to the cgroup relative path, which seems 
> > wrong to me.
> >     
> >     2. Is it expected that there's always a parent in this call? Otherwise 
> > seems

First, every hierarchy has a "root" cgroup at 'hierarchy' (i.e., 
'path::join(hierarchy, "/")'). So by definition, yes, every cgroup has a parent.

That being said, someone could be creating a cgroup called "foo" or "/foo" 
(since we allow both right now, even though we specify that we want a relative 
path). In the former case (the "good" relative path), I needed to be able to 
determine the parent cgroup so I know what cgroup to clone from. Simply calling 
dirname on this should return '.', which I could make work, but that would 
require updating a lot of code which doesn't currently call os::realpath to do 
so. I decided to save that for another day.


> On Oct. 18, 2012, 5:09 p.m., Ben Mahler wrote:
> > third_party/libprocess/include/stout/path.hpp, line 19
> > <https://reviews.apache.org/r/7620/diff/1/?file=177250#file177250line19>
> >
> >     Thanks for this, I seem to recall making some clunky join calls that I 
> > should fix.

Yup.


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/7620/#review12573
-----------------------------------------------------------


On Oct. 24, 2012, 4:40 a.m., Benjamin Hindman wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/7620/
> -----------------------------------------------------------
> 
> (Updated Oct. 24, 2012, 4:40 a.m.)
> 
> 
> Review request for mesos, Vinod Kone, Jie Yu, and Ben Mahler.
> 
> 
> Description
> -------
> 
> See summary.
> 
> 
> Diffs
> -----
> 
>   src/linux/cgroups.hpp f0e343d9348075484fa5a4be739a5cef8b9dea21 
>   src/linux/cgroups.cpp 58e9fec193cf0da39fdbee3ee1891fd716a63bbb 
>   third_party/libprocess/include/stout/path.hpp 
> e64f8a01d004c5e6513880ed2a1a85d4ec4f9e8b 
> 
> Diff: https://reviews.apache.org/r/7620/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Benjamin Hindman
> 
>

Reply via email to