Date: Wed, 31 Dec 2003 17:32:49 +0200 From: Valentin Nechayev <netch@netch.kiev.ua> To: Poul-Henning Kamp <phk@phk.freebsd.dk> Cc: Valentin Nechayev <netch@netch.kiev.ua> Subject: Re: geom bsd strangeness in extended partition Message-ID: <20031231153249.GA434@iv.nn.kiev.ua> In-Reply-To: <18000.1072741953@critter.freebsd.dk> References: <20031229230048.GA310@iv.nn.kiev.ua> <18000.1072741953@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
Tue, Dec 30, 2003 at 00:52:33, phk (Poul-Henning Kamp) wrote about "Re: geom bsd strangeness in extended partition":
>>GEOM issues warnings during scanning bsdlabels in extended partition:
PHK> This is a bug and I know what it is. I'll get back to you with
PHK> a patch sometimes in january.
Well, the attached patch is seems to be dirty, but fixed it for me.
I don't understand why multiple customers are allowed for g_geom, so get
one of them which leads to parent MBR.
Some variables may be unused because I omit debug print from final patch.
-netch-
--- geom/geom_mbr.c Mon Sep 1 23:45:32 2003
+++ geom.1/geom_mbr.c Wed Dec 31 17:22:10 2003
@@ -299,6 +299,29 @@
mp = gsp->softc;
if (bp->bio_cmd == BIO_GETATTR) {
if (g_handleattr_int(bp, "MBR::type", mp->type[idx]))
+ return (1);
+ off_t ro, off2;
+ struct g_consumer *pcons;
+ struct g_provider *pp2;
+ struct g_geom *gp2;
+ struct g_slicer *gsp2;
+ //struct g_mbr_softc *mp2;
+ /* Add MBR::offset of MBR in parent geom */
+ ro = gsp->slices[idx].offset;
+ LIST_FOREACH(pcons, &gp->consumer, consumer) {
+ if (!(pp2 = pcons->provider) ||
+ !(gp2 = pp2->geom) ||
+ !gp2->class ||
+ strcmp(gp2->class->name, MBR_CLASS_NAME))
+ continue;
+ gsp2 = gp2->softc;
+ off2 = gsp2->slices[pp2->index].offset;
+ ro += off2;
+ break;
+ }
+ if (g_handleattr_off_t(bp, "MBR::offset", ro))
+ return (1);
+ if (g_handleattr_off_t(bp, "MBR::offset", ro))
return (1);
}
return (0);
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20031231153249.GA434>