On Thu, 2008-08-07 at 18:20 +0200, Michael Kerrisk wrote: > On Thu, Aug 7, 2008 at 5:27 PM, Lee Schermerhorn > <Lee.Schermerhorn@xxxxxx> wrote: > > On Thu, 2008-08-07 at 15:33 +0200, Michael Kerrisk wrote: > >> Lee, > >> > >> Lee Schermerhorn wrote: > >> > Misc cleanup of get_mempolicy(2): > >> > > >> > + mention that any mode flags will be saved with mode. > >> > I don't bother to document mode flags here because we > >> > already have a pointer to set_mempolicy(2) for more info > >> > on memory policy. mode flags are discussed there. > >> > > >> > + remove some old, obsolete [IMO] NOTES > >> > >> Can you say a little more about the rationale for removing that piece of > >> text. (I'm not objecting: I'd just like al little more context to the change.) > > > > 1) no longer true that we have no cpuset man page to refer to. > > > > 2) We documented MPOL_F_NODE several man page revisions ago. So the old > > comment that it's not documented, and my rationale for documenting > > it--effectively, Christoph and Andi agreed--are both old news. The > > changes should be captured in your change logs, so I think the in-line > > comments can be removed. > > Sorry -- maybe I should have been clearer. I meant, what's the > rationale for removing this piece of text: > > -If the mode of the process policy or the policy governing allocations at the > -specified address is > -.B MPOL_PREFERRED > -and this policy was installed with an empty > -.IR nodemask \(emspecifying > -local allocation, > -.BR get_mempolicy () > -will return the mask of on-line node IDs in the location pointed to by > -a non-NULL > -.I nodemask > -argument. > -This mask does not take into consideration any administratively imposed > -restrictions on the process's context. Ah, yes. Well, it doesn't do that. get_mempolicy() always returns the mempolicy that can be used by set_mempolicy() [or mbind()], to establish the same policy. So, if one specified MPOL_PREFERRED with a NULL or empty nodemask, that's what get_mempolicy() returns [even tho' it's stored differently internally]. > > Cheers, > > Michael > > > > >> > >> Cheers, > >> > >> Michael > >> > >> > and 'roff comments. > >> > > >> > Signed-off-by: Lee Schermerhorn <lee.schermerhorn@xxxxxx> > >> > > >> > man2/get_mempolicy.2 | 44 +++++++++++++------------------------------- > >> > 1 files changed, 13 insertions(+), 31 deletions(-) > >> > > >> > Index: man-pages-3.05/man2/get_mempolicy.2 > >> > =================================================================== > >> > --- man-pages-3.05.orig/man2/get_mempolicy.2 2008-07-29 16:43:07.000000000 -0400 > >> > +++ man-pages-3.05/man2/get_mempolicy.2 2008-07-29 16:43:40.000000000 -0400 > >> > @@ -53,6 +53,15 @@ then information about the calling proce > >> > (as set by > >> > .BR set_mempolicy (2)) > >> > is returned. > >> > +The policy returned > >> > +.RI [ mode > >> > +and > >> > +.IR nodemask ] > >> > +may be used to restore the process' policy to it's state at > >> > +the time of the call to > >> > +.BR get_mempolicy () > >> > +using > >> > +.BR set_mempolicy (2). > >> > > >> > If > >> > .I flags > >> > @@ -72,7 +81,10 @@ If the > >> > .I mode > >> > argument is not NULL, then > >> > .BR get_mempolicy () > >> > -will store the policy mode of the requested NUMA policy in the location > >> > +will store the policy mode > >> > +and any optional > >> > +.I "mode flags" > >> > +of the requested NUMA policy in the location > >> > pointed to by this argument. > >> > If > >> > .I nodemask > >> > @@ -188,36 +200,6 @@ cases.) > >> > .SH CONFORMING TO > >> > This system call is Linux-specific. > >> > .SH NOTES > >> > -If the mode of the process policy or the policy governing allocations at the > >> > -specified address is > >> > -.B MPOL_PREFERRED > >> > -and this policy was installed with an empty > >> > -.IR nodemask \(emspecifying > >> > -local allocation, > >> > -.BR get_mempolicy () > >> > -will return the mask of on-line node IDs in the location pointed to by > >> > -a non-NULL > >> > -.I nodemask > >> > -argument. > >> > -This mask does not take into consideration any administratively imposed > >> > -restrictions on the process's context. > >> > -.\" FIXME > >> > -.\" "context" above refers to cpusets. > >> > -.\" No man page to reference. --Lee Schermerhorn > >> > -.\" > >> > -.\" Christoph says the following is untrue. These are "fully supported." > >> > -.\" Andi concedes that he has lost this battle and approves [?] > >> > -.\" updating the man pages to document the behavior. --Lee Schermerhorn > >> > -.\" This manual page is incomplete: > >> > -.\" it does not document the details the > >> > -.\" .BR MPOL_F_NODE > >> > -.\" flag, > >> > -.\" which modifies the operation of > >> > -.\" .BR get_mempolicy (). > >> > -.\" This is deliberate: this flag is not intended for application use, > >> > -.\" and its operation may change or it may be removed altogether in > >> > -.\" future kernel versions. > >> > -.\" .B Do not use it. > >> > .SS "Versions and Library Support" > >> > See > >> > .BR mbind (2). > >> > > >> > > > > > > > -- To unsubscribe from this list: send the line "unsubscribe linux-numa" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
