Date: Thu, 21 Nov 2013 22:02:59 +0000 (UTC) From: Marcel Moolenaar <marcel@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r258448 - head/sys/geom/part Message-ID: <201311212202.rALM2xAS078472@svn.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: marcel Date: Thu Nov 21 22:02:59 2013 New Revision: 258448 URL: http://svnweb.freebsd.org/changeset/base/258448 Log: Have the GPT probe return a lower priority when the MBR is not a PMBR The purpose of the PMBR is to have the disk appear in use to GPT unaware utilities (like fdisk). However, if the PMBR has been changed by a GPT unaware utlity then we must assume that this was deliberate (as it involved removal of the special slice) and we should not treat the unmodified GPT-specific sectors as being valid. By lowering the probe priority in that case, the MBR scheme will take precedence and the kernel will end up using the MBR and not the GPT. We will still use the GPT if the kernel does not support the MBR scheme. Modified: head/sys/geom/part/g_part_gpt.c Modified: head/sys/geom/part/g_part_gpt.c ============================================================================== --- head/sys/geom/part/g_part_gpt.c Thu Nov 21 21:25:58 2013 (r258447) +++ head/sys/geom/part/g_part_gpt.c Thu Nov 21 22:02:59 2013 (r258448) @@ -759,8 +759,8 @@ static int g_part_gpt_probe(struct g_part_table *table, struct g_consumer *cp) { struct g_provider *pp; - char *buf; - int error, res; + u_char *buf; + int error, index, pri, res; /* We don't nest, which means that our depth should be 0. */ if (table->gpt_depth != 0) @@ -785,11 +785,21 @@ g_part_gpt_probe(struct g_part_table *ta if (pp->sectorsize < MBRSIZE || pp->mediasize < 6 * pp->sectorsize) return (ENOSPC); - /* Check that there's a MBR. */ + /* + * Check that there's a MBR or a PMBR. If it's a PMBR, we return + * as the highest priority on a match, otherwise we assume some + * GPT-unaware tool has destroyed the GPT by recreating a MBR and + * we really want the MBR scheme to take precedence. + */ buf = g_read_data(cp, 0L, pp->sectorsize, &error); if (buf == NULL) return (error); res = le16dec(buf + DOSMAGICOFFSET); + pri = G_PART_PROBE_PRI_LOW; + for (index = 0; index < NDOSPART; index++) { + if (buf[DOSPARTOFF + DOSPARTSIZE * index + 4] == 0xee) + pri = G_PART_PROBE_PRI_HIGH; + } g_free(buf); if (res != DOSMAGIC) return (ENXIO); @@ -801,7 +811,7 @@ g_part_gpt_probe(struct g_part_table *ta res = memcmp(buf, GPT_HDR_SIG, 8); g_free(buf); if (res == 0) - return (G_PART_PROBE_PRI_HIGH); + return (pri); /* No primary? Check that there's a secondary. */ buf = g_read_data(cp, pp->mediasize - pp->sectorsize, pp->sectorsize, @@ -810,7 +820,7 @@ g_part_gpt_probe(struct g_part_table *ta return (error); res = memcmp(buf, GPT_HDR_SIG, 8); g_free(buf); - return ((res == 0) ? G_PART_PROBE_PRI_HIGH : ENXIO); + return ((res == 0) ? pri : ENXIO); } static int
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201311212202.rALM2xAS078472>