----- Original Message -----
> From: chru...@suse.cz
> To: "Jan Stancek" <jstan...@redhat.com>
> Cc: ltp-list@lists.sourceforge.net
> Sent: Tuesday, 29 July, 2014 10:43:06 AM
> Subject: Re: [LTP] [PATCH] mem/oom: check for multiple outcomes if
> overcommit_memory is 0 or 1
>
> Hi!
> > +/* oom/testoom outcomes */
> > +#define EXIT_FLAG 0x100
> > +#define SIGNAL_FLAG 0x200
> > +#define OUTCOME_MASK 0xff
> > +#define EXITED(n) (EXIT_FLAG + (n & OUTCOME_MASK))
> > +#define SIGNALLED(n) (SIGNAL_FLAG + (n & OUTCOME_MASK))
> > +
> > +extern int EXITED_ENOMEM[];
> > +extern int ENOMEM_OR_SIGKILL[];
>
> This machinery looks a bit too generic to me. IMHO simple bit flag for
> oom() that would allow/disallow the child being killed as a PASS result
> would suffice.
>
> Or do you expect that we will need (many) more possible cases to add?
Not many. ENOMEM/SIGKILL or 0 (meaning child completed without issues)
are the most likely ones we'll need.
How about:
oom(int testcase, int lite, int retcode, int can_be_killed)
Regards,
Jan
>
> > long overcommit;
> > -void oom(int testcase, int lite);
> > -void testoom(int mempolicy, int lite);
> > +void oom(int testcase, int lite, int outcome[]);
> > +void testoom(int mempolicy, int lite, int outcome[]);
> >
> > /* KSM */
> >
> > diff --git a/testcases/kernel/mem/lib/mem.c
> > b/testcases/kernel/mem/lib/mem.c
> > index c7910db..f207c58 100644
> > --- a/testcases/kernel/mem/lib/mem.c
> > +++ b/testcases/kernel/mem/lib/mem.c
> > @@ -25,52 +25,73 @@
> > #include "numa_helper.h"
> >
> > /* OOM */
> > +int EXITED_ENOMEM[] = { EXITED(ENOMEM), 0 };
> > +int ENOMEM_OR_SIGKILL[] = { EXITED(ENOMEM), SIGNALLED(SIGKILL), 0 };
> >
> > static int alloc_mem(long int length, int testcase)
> > {
> > - void *s;
> > + char *s;
> > + long i;
> > + int pagesz = getpagesize();
> >
> > tst_resm(TINFO, "allocating %ld bytes.", length);
> > +
> > s = mmap(NULL, length, PROT_READ | PROT_WRITE,
> > MAP_ANONYMOUS | MAP_PRIVATE, -1, 0);
> > - if (s == MAP_FAILED) {
> > - if (testcase == OVERCOMMIT && errno == ENOMEM)
> > - return 1;
> > - else
> > - tst_brkm(TBROK | TERRNO, cleanup, "mmap");
> > - }
> > + if (s == MAP_FAILED)
> > + return errno;
> > +
> > if (testcase == MLOCK && mlock(s, length) == -1)
> > - tst_brkm(TBROK | TERRNO, cleanup, "mlock");
> > + return errno;
> > #ifdef HAVE_MADV_MERGEABLE
> > if (testcase == KSM && madvise(s, length, MADV_MERGEABLE) == -1)
> > - tst_brkm(TBROK | TERRNO, cleanup, "madvise");
> > + return errno;
> > #endif
> > - memset(s, '\a', length);
> > + for (i = 0; i < length; i += pagesz)
> > + s[i] = '\a';
>
> This is good optimalization but should be either mentioned in the commit
> message or ideally commited separately.
>
> --
> Cyril Hrubis
> chru...@suse.cz
>
------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list
