Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 5 Nov 2017 11:42:51 -0700
From:      Warner Losh <imp@bsdimp.com>
To:        "Conrad E. Meyer" <cem@freebsd.org>
Cc:        Konstantin Belousov <kostikbel@gmail.com>, src-committers <src-committers@freebsd.org>,  "svn-src-all@freebsd.org" <svn-src-all@freebsd.org>,  "svn-src-head@freebsd.org" <svn-src-head@freebsd.org>
Subject:   Re: svn commit: r325386 - head/sys/kern
Message-ID:  <CANCZdfqApxd99G-HqVfe4x8EysYJVCAcHksqTeUuBAaYR1OjEA@mail.gmail.com>
In-Reply-To: <CAG6CVpVQ4RSgz883ezsdExVDQ4RbSiR_y%2BZ5XWDNFeYBpMS-cQ@mail.gmail.com>
References:  <201711041049.vA4AnZUE096709@repo.freebsd.org> <CAG6CVpUaXSX26Bc839kn96EXnUjtGyQ3_eNJPhsRB%2Bv6G6gH1Q@mail.gmail.com> <20171105130607.GA2566@kib.kiev.ua> <CAG6CVpV3SfV8VuNeMJEuN%2BMSu3424mK=HO_-YW9vRt9HEdcJZQ@mail.gmail.com> <20171105173032.GE2566@kib.kiev.ua> <CAG6CVpVQ4RSgz883ezsdExVDQ4RbSiR_y%2BZ5XWDNFeYBpMS-cQ@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Nov 5, 2017 at 11:32 AM, Conrad Meyer <cem@freebsd.org> wrote:

> E.g.,
>
> --- a/sys/ufs/ffs/ffs_alloc.c
> +++ b/sys/ufs/ffs/ffs_alloc.c
> @@ -304,8 +304,7 @@ retry:
>         }
>
>         if (bp->b_blkno == bp->b_lblkno) {
> -               if (lbprev >= UFS_NDADDR)
> -                       panic("ffs_realloccg: lbprev out of range");
> +               ASSERT(lbprev < UFS_NDADDR, "ffs_realloccg: lbprev out
> of range");
>                 bp->b_blkno = fsbtodb(fs, bprev);
>         }
>

Just a side point: All these should be programming errors. The bogus data
that comes or could come from the FS itself should remain always-on panics.
Well, actually, they should transition from always-on panics to some sort
of degraded mount that would be more resilient in the face of such
corruption. But failing that, they should remain always-on panics :)

Warner



> On Sun, Nov 5, 2017 at 9:30 AM, Konstantin Belousov <kostikbel@gmail.com>
> wrote:
> > On Sun, Nov 05, 2017 at 09:16:28AM -0800, Conrad Meyer wrote:
> >> On Sun, Nov 5, 2017 at 5:06 AM, Konstantin Belousov <
> kostikbel@gmail.com> wrote:
> >> > On Sat, Nov 04, 2017 at 12:04:56PM -0700, Conrad Meyer wrote:
> >> >> This is a functional change, because MPASS (via KASSERT) is only
> >> >> enabled on DEBUG kernels.  Ideally we would have a kind of ASSERT
> that
> >> >> worked on NODEBUG kernels.
> >> > Why would we need such thing ?
> >> >
> >> > Our conventions are clear: consistency checks are normally done with
> >> > KASSERT() and enabled for DEBUG (INVARIANTS or harder) configurations.
> >> > We only leave explicit panics in the production kernels when there
> >> > continuation of operations is worse then abort, e.g. when UFS detects
> >> > the metadata corruption.
> >>
> >> An always-on assert construct would be precisely for the latter
> >> scenario.  Instead, we litter the tree with "if (!invariant) {
> >> panic(); }."
> > We do
> >
> > #ifdef INVARIANTS
> > if (!condition) panic();
> > #endif
> >
> > I do not understand what do you mean by 'instead'.
>
>



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CANCZdfqApxd99G-HqVfe4x8EysYJVCAcHksqTeUuBAaYR1OjEA>