Date: Thu, 04 Mar 2004 16:21:29 -0800 From: Cy Schubert <Cy.Schubert@komquats.com> To: Don Lewis <truckman@freebsd.org> Cc: freebsd-current@freebsd.org Subject: Re: sysctl spinning (was: Re: ps Causes Hard Hang) Message-ID: <200403050021.i250LU0f015396@cwsys.cwsent.com> In-Reply-To: Your message of "Thu, 04 Mar 2004 13:42:28 PST." <200403042142.i24LgS7E005323@gw.catspoiler.org>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <200403042142.i24LgS7E005323@gw.catspoiler.org>, Don Lewis writes: > On 5 Mar, Bruce Evans wrote: > > On Thu, 4 Mar 2004, Don Lewis wrote: > > > >> I just checked the mlock() man page in the Single UNIX Specification > >> Version 3. Our mlock() implementation is broken. We should be > >> returning ENOMEM here, though this will result in some sort of user > >> visible sysctl breakage instead of a tight loop. > > > > POSIX is more authoritative. It says that ENOMEM may be returned if > > an implementation-defined limit would be exceeded, and that EAGAIN > > shall be returned if some or all of the memory could not be locked > > when the call was made. So it literally requires returning EAGAIN > > even when the limit would be exceeded, but it doesn't mean to require > > that. > > That's the same as SUSv3. > > I was pretty frazzled when I sent my previous message and was thinking > the test was something like: > if (atop(size) > vm_page_max_wired) > > The following patch is probably better. > > Index: sys/vm/vm_mmap.c > =================================================================== > RCS file: /home/ncvs/src/sys/vm/vm_mmap.c,v > retrieving revision 1.181 > diff -u -r1.181 vm_mmap.c > --- sys/vm/vm_mmap.c 1 Mar 2004 02:44:33 -0000 1.181 > +++ sys/vm/vm_mmap.c 4 Mar 2004 21:24:29 -0000 > @@ -923,8 +923,8 @@ > if (addr + size < addr) > return (EINVAL); > > - if (atop(size) + cnt.v_wire_count > vm_page_max_wired) > - return (EAGAIN); > + if (atop(size) > vm_page_max_wired) > + return (ENOMEM); > > PROC_LOCK(proc); > if (size + ptoa(pmap_wired_count(vm_map_pmap(&proc->p_vmspace->vm_map)) > ) > > @@ -933,6 +933,9 @@ > return (ENOMEM); > } > PROC_UNLOCK(proc); > + > + if (atop(size) + cnt.v_wire_count > vm_page_max_wired) > + return (EAGAIN); > > error = vm_map_wire(&proc->p_vmspace->vm_map, addr, > addr + size, VM_MAP_WIRE_USER|VM_MAP_WIRE_NOHOLES); > I'll give this a try tonight after my buildworld completes. I'll keep you posted. Cheers, -- Cy Schubert <Cy.Schubert@komquats.com> http://www.komquats.com/ BC Government . FreeBSD UNIX Cy.Schubert@osg.gov.bc.ca . cy@FreeBSD.org http://www.gov.bc.ca/ . http://www.FreeBSD.org/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200403050021.i250LU0f015396>