On Tue, Aug 03, 2004 at 07:58:11PM +0200, Luca Berra wrote:
> hi,
> i am testing dmraid 1.0.0-rc2 on Mandrakelinux cooker
>
> i have some issue building it, see the attached patch.
> 1) le32_to_cpu and le16_to_cpu are protected by an #ifdef __KERNEL__ on
> my glibc version, i redefine them in utils.h, since that file does not
> exist :)
Hrm, not here.
Have created a datastruct/byteorder.h instead to stay away from
this dependency.
> 2) hd_big_geometry has been removed from 2.6
Changed.
> 3) better check the fs_type to find sysfs, since the device name can be
> an arbitrary string
Changed.
> 4) new versions of strip fail to strip read only files, could you avoid
> setting the file readonly?
Changed.
>
> besides that i have some weird behaviour with your makefile
> (maybe a make bug on my side, but i never saw it)
> if i use make -s
> or
> make -d
> it builds correctly
> if i use make without options it continuously loops on the generation of
> .d files.
Same here with make 3.80. 3.79 version does it just fine.
Any idea why that bogus behaviour gets triggered in make 3.80 is welcome.
>
>
> last,
> it looks like that when i try to partition my hpt 37x (raid 1) from dos
> it creates a partition starting at sector 1 on the physical disks, not
> at sector 10, could this be possible?
Yes, the first drive in a hpt37x set gets mapped at sector 0, not 10.
Firts partition should start at a later sector than one though.
Fix is in my code base already and has been tested successfully.
All changes vailable in 1.0.0-rc3 later this week...
Regards,
Heinz -- The LVM Guy --
>
> regards,
> L.
>
>
> --
> Luca Berra -- bluca@xxxxxxxxxx
> Communication Media & Services S.r.l.
> /"\
> \ / ASCII RIBBON CAMPAIGN
> X AGAINST HTML MAIL
> / \
> --- dmraid/1.0.0-rc2/include/utils.h.mdk 2004-07-24 12:01:26.000000000 +0200
> +++ dmraid/1.0.0-rc2/include/utils.h 2004-07-24 12:01:26.000000000 +0200
> @@ -0,0 +1,3 @@
> +#include <asm/byteorder.h>
> +#define le32_to_cpu __le32_to_cpu
> +#define le16_to_cpu __le16_to_cpu
> --- dmraid/1.0.0-rc2/lib/device/dev-io.h.mdk 2004-07-15 16:10:59.000000000 +0200
> +++ dmraid/1.0.0-rc2/lib/device/dev-io.h 2004-07-24 12:01:26.000000000 +0200
> @@ -38,7 +38,11 @@
> char *path;
> char *serial; /* SCSI serial number. */
> struct dmraid_dev *rd;
> +#ifdef HDIO_GETGEO_BIG
> struct hd_big_geometry geometry;
> +#else
> + struct hd_geometry geometry;
> +#endif
> uint64_t sectors;
> };
>
> --- dmraid/1.0.0-rc2/lib/device/scan.c.mdk 2004-07-15 21:24:52.000000000 +0200
> +++ dmraid/1.0.0-rc2/lib/device/scan.c 2004-07-24 12:03:14.000000000 +0200
> @@ -42,7 +42,7 @@
> LOG_ERR(0, "Failed to open /etc/mtab");
>
> while ((ent = getmntent(mtab))) {
> - if (!strcmp(ent->mnt_fsname, "sysfs")) {
> + if (!strcmp(ent->mnt_type, "sysfs")) {
> ret = ent->mnt_dir;
> break;
> }
> --- dmraid/1.0.0-rc2/tools/Makefile.in.mdk 2004-07-24 12:01:26.000000000 +0200
> +++ dmraid/1.0.0-rc2/tools/Makefile.in 2004-07-24 12:01:26.000000000 +0200
> @@ -59,7 +59,7 @@
>
> install_dmraid_tools: $(TARGETS)
> @echo "Installing $(TARGETS) in $(sbindir)"
> - $(INSTALL) -o $(OWNER) -g $(GROUP) -m 555 $(STRIP) $(TARGETS) \
> + $(INSTALL) -o $(OWNER) -g $(GROUP) $(STRIP) $(TARGETS) \
> $(sbindir)
>
> install: install_dmraid_tools
> _______________________________________________
>
> Ataraid-list@xxxxxxxxxx
> https://www.redhat.com/mailman/listinfo/ataraid-list
*** Software bugs are stupid.
Nevertheless it needs not so stupid people to solve them ***
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Heinz Mauelshagen Red Hat GmbH
Consulting Development Engineer Am Sonnenhang 11
56242 Marienrachdorf
Germany
Mauelshagen@xxxxxxxxxx +49 2626 141200
FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
