On Monday 08 December 2014 10:18:01 Jan Beulich wrote:
> >>> On 08.12.14 at 03:30, <mdo...@bitdefender.com> wrote:
> > +#ifndef NDEBUG
> > +static bool_t xmem_pool_check_size(const struct bhdr *b, int fl, int sl)
> > +{
> > + while ( b )
> > + {
> > + int __fl;
> > + int __sl;
> > +
> > + MAPPING_INSERT(b->size, &__fl, &__sl);
> > + if ( __fl != fl || __sl != sl )
> > + {
> > + printk(XENLOG_ERR "xmem_pool: for block %p size = %u, { fl =
> > %d, sl = %d } should be { fl = %d, sl = %d }\n",
>
> Quoting my reply to v1: "Long line. Only the format message alone
> is allowed to exceed 80 characters."
>
Just so I don't send another faulty patch, you would see that printk()
being:
printk(XENLOG_ERR
"xmem_pool: for block %p size = %u, { fl = %d, sl = %d } should be {
fl = %d, sl = %d }\n",
b, b->size, fl, sl, __fl, __sl);
?
> Also with there potentially being multiple pools, shouldn't all of the
> log messages the patch issues be extended to allow identifying the
> offending one?
>
I think I can insert the pool name in that message too. Something like:
printk(XENLOG_ERR
"xmem_pool: %s: for block [...]\n",
pool->name, b, b->size [...]);
would do? A quick preview:
[2014-12-04 15:41:23] (XEN) [ 1374.507125] xmem_pool: xmalloc: for block
ffff8304004fb9b0 size = 0, { fl = 3, sl = 9 } should be { fl = 0, sl = 0 }
[2014-12-04 15:41:23] (XEN) [ 1374.507127] xmem_pool: xmalloc: the TLSF chunk
matrix is corrupted
> > +bool_t __xmem_pool_check(const char *file, int line, struct xmem_pool
> > *pool)
> > +{
> > + return __xmem_pool_check_unlocked(file, line, pool ? pool : xenpool);
>
> For brevity, the shorter "pool ?: xenpool" is generally preferable. The
> only place using this is not allowed are the public headers.
>
Will do.
Thank you,
--
Mihai DONȚU
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
