I am using the srpm from http://xenbits.xensource.com/kernels/rhel3x/kernel-2.4.21-47.0.1.EL.xs0.3.5. 15.src.rpm (I get the same issue using the binary RPM) the dom0 is running 3.0.3_0 Upon booting the DomU, (the DomU has been passed phy:/dev/sda6, which has been partitioned using qemu) I get the following ioctl errors. ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif relivant bit from dmesg follows: Freeing initrd memory: 8000k freed VFS: Mounted root (ext2 filesystem). Journalled Block Device driver loaded kjournald starting. Commit interval 5 seconds EXT3-fs: mounted filesystem with ordered data mode. Freeing unused kernel memory: 136k freed EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,2), internal journal Adding Swap: 1044216k swap-space (priority -1) ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif ioctl 0000126c not supported by XL blkif kjournald starting. Commit interval 5 seconds EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,1), internal journal EXT3-fs: mounted filesystem with ordered data mode. kjournald starting. Commit interval 5 seconds EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,6), internal journal EXT3-fs: mounted filesystem with ordered data mode. kjournald starting. Commit interval 5 seconds EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,5), internal journal EXT3-fs: mounted filesystem with ordered data mode. (I asked this question last week on Xen-users with less information and got no answers, I''m still stuck on it; Sorry for the re-post. If someone could point me in the correct direction, I would be extremely grateful.) _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Ian Campbell
2006-Dec-06 10:41 UTC
Re: [Xen-devel] ioctl 0000126c not supported by XL blkif
On Tue, 2006-12-05 at 10:44 -0800, Luke Crawford wrote:> I am using the srpm from > http://xenbits.xensource.com/kernels/rhel3x/kernel-2.4.21-47.0.1.EL.xs0.3.5. > 15.src.rpm > > (I get the same issue using the binary RPM) > > the dom0 is running 3.0.3_0 > > Upon booting the DomU, (the DomU has been passed phy:/dev/sda6, which has > been partitioned using qemu) I get the following ioctl errors. > > ioctl 0000126c not supported by XL blkif0x12 seems to be the type used for the BLK* ioctls defined in include/linux/fs.h and 0x6c==108 is listed as BLKGETLASTSECT. Strangely it is #ifdef CONFIG_IA64 everywhere in the kernel source (header and drivers/block/blkpg.c). In the system wide version (/usr/include/linux/fs.h) it is defined unconditionally. There is also a comment in the source which says "108-111 have been used for various private purposes". The IOCTL was added by linux-2.4.21-lastsector.patch from the SRPM. Anyway, I don''t see these errors on my RHEL38 guest during normal booting, only when I do something extra like running "parted". You are running a RHEL3x userspace aren''t you? Could you try the patch below, hopefully it will give some indication which process is calling the unimplemented ioctl. Once we have confirmed for sure that it is BLKGETLASTSECT on x86 we can look at fixing it up. My feeling is that the warning is pretty harmless. Though it might bite you if you wanted to repartition or something. Ian. [BLKFRONT] Report the process making an unsupported IOCTL. Signed-off-by: Ian Campbell <ian.campbell@xensource.com> diff -r d2d1872fce96 drivers/xen/blkfront/blkfront.c --- a/drivers/xen/blkfront/blkfront.c Mon Dec 04 09:32:41 2006 +0000 +++ b/drivers/xen/blkfront/blkfront.c Wed Dec 06 10:31:31 2006 +0000 @@ -621,7 +621,8 @@ int blkif_ioctl(struct inode *inode, str return -ENOSYS; default: - printk(KERN_ALERT "ioctl %08x not supported by XL blkif\n", command); + printk(KERN_ALERT "ioctl %08x from pid %d \"%s\" not supported by XL blkif\n", + command, current->pid, current->comm); return -ENOSYS; } _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Luke Crawford
2006-Dec-07 02:44 UTC
Re: [Xen-devel] ioctl 0000126c not supported by XL blkif
On Wed, 6 Dec 2006, Ian Campbell wrote:> Anyway, I don''t see these errors on my RHEL38 guest during normal > booting, only when I do something extra like running "parted". You are > running a RHEL3x userspace aren''t you?Yes, I am.> Could you try the patch below, hopefully it will give some indication > which process is calling the unimplemented ioctl. Once we have confirmed > for sure that it is BLKGETLASTSECT on x86 we can look at fixing it up.ioctl 0000126c from pid 130 "partx" not supported by XL blkif ioctl 0000126c from pid 145 "partx" not supported by XL blkif ioctl 0000126c from pid 162 "partx" not supported by XL blkif ioctl 0000126c from pid 179 "partx" not supported by XL blkif ioctl 0000126c from pid 194 "partx" not supported by XL blkif ioctl 0000126c from pid 209 "partx" not supported by XL blkif> My feeling is that the warning is pretty harmless. Though it might bite > you if you wanted to repartition or something.partx. I guess I need to go do some googling on why that is there. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Ian Campbell
2006-Dec-07 09:45 UTC
Re: [Xen-devel] ioctl 0000126c not supported by XL blkif
On Wed, 2006-12-06 at 18:44 -0800, Luke Crawford wrote:> On Wed, 6 Dec 2006, Ian Campbell wrote:> > Could you try the patch below, hopefully it will give some indication > > which process is calling the unimplemented ioctl. Once we have confirmed > > for sure that it is BLKGETLASTSECT on x86 we can look at fixing it up. > > ioctl 0000126c from pid 130 "partx" not supported by XL blkif > ioctl 0000126c from pid 145 "partx" not supported by XL blkif > ioctl 0000126c from pid 162 "partx" not supported by XL blkif > ioctl 0000126c from pid 179 "partx" not supported by XL blkif > ioctl 0000126c from pid 194 "partx" not supported by XL blkif > ioctl 0000126c from pid 209 "partx" not supported by XL blkif > > > My feeling is that the warning is pretty harmless. Though it might bite > > you if you wanted to repartition or something. > > partx. I guess I need to go do some googling on why that is there.It comes from the util-linux package. partx only makes this ioctl call if you have an odd number of sectors in your diskimage, and only when trying to process an "EFI GUID Partition Table" (whatever that is, I presume it is an IA64 thing, although the partx code is unconditional). I assume you aren''t really using this type of partition table so I expect it will fail harmlessly. Do you have a disk setup which might cause it to think it needs to try the more unusual partition types? Perhaps you are using an unpartitioned disk of an odd size? That reproduced it for me. Ian. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Luke Crawford
2006-Dec-07 19:15 UTC
Re: [Xen-devel] ioctl 0000126c not supported by XL blkif
On Thu, 7 Dec 2006, Ian Campbell wrote:> > partx only makes this ioctl call if you have an odd number of sectors in > your diskimage, and only when trying to process an "EFI GUID Partition > Table" (whatever that is, I presume it is an IA64 thing, although the > partx code is unconditional).I see lots of references to "odd" partition sizes in my google, but I''m missing what ''odd'' means in this context. 0dd as opposed to even? or odd as in "not aligning with cyl boundaries" or something like that? (I suppose my partitions listed below are odd in both respects)> I assume you aren''t really using this type of partition table so I > expect it will fail harmlessly. > > Do you have a disk setup which might cause it to think it needs to try > the more unusual partition types? Perhaps you are using an unpartitioned > disk of an odd size? That reproduced it for me.we are using a systemimager floppy booted off qemu to create the DomU, so this is the same partition layout that would be given to a physical box. [root@DomU]# fdisk -l /dev/sda Disk /dev/sda: 36.5 GB, 36531973632 bytes 255 heads, 63 sectors/track, 4441 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/sda1 * 1 16 128488+ 83 Linux /dev/sda2 17 669 5245222+ 83 Linux /dev/sda3 670 799 1044225 83 Linux /dev/sda4 800 4441 29254365 f Win95 Ext''d (LBA) /dev/sda5 800 930 1052226 83 Linux /dev/sda6 931 4441 28202076 83 Linux _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Ian Campbell
2006-Dec-07 20:39 UTC
Re: [Xen-devel] ioctl 0000126c not supported by XL blkif
On Thu, 2006-12-07 at 11:15 -0800, Luke Crawford wrote:> On Thu, 7 Dec 2006, Ian Campbell wrote: > > > > partx only makes this ioctl call if you have an odd number of sectors in > > your diskimage, and only when trying to process an "EFI GUID Partition > > Table" (whatever that is, I presume it is an IA64 thing, although the > > partx code is unconditional). > > I see lots of references to "odd" partition sizes in my google, but I''m > missing what ''odd'' means in this context. 0dd as opposed to even? or > odd as in "not aligning with cyl boundaries" or something like that? > (I suppose my partitions listed below are odd in both respects)Odd as opposed to even number of sectors.> > I assume you aren''t really using this type of partition table so I > > expect it will fail harmlessly. > > > > Do you have a disk setup which might cause it to think it needs to try > > the more unusual partition types? Perhaps you are using an unpartitioned > > disk of an odd size? That reproduced it for me. > > we are using a systemimager floppy booted off qemu to create the DomU, so > this is the same partition layout that would be given to a physical box. > > [root@DomU]# fdisk -l /dev/sda > > Disk /dev/sda: 36.5 GB, 36531973632 bytes > 255 heads, 63 sectors/track, 4441 cylinders > Units = cylinders of 16065 * 512 = 8225280 bytes > > Device Boot Start End Blocks Id System > /dev/sda1 * 1 16 128488+ 83 Linux > /dev/sda2 17 669 5245222+ 83 Linux > /dev/sda3 670 799 1044225 83 Linux > /dev/sda4 800 4441 29254365 f Win95 Ext''d (LBA) > /dev/sda5 800 930 1052226 83 Linux > /dev/sda6 931 4441 28202076 83 LinuxThat would do it. I''m going to remove the message tomorrow -- all of the other kernel ports have it commented out already. It was only really useful during development anyway. Cheers, Ian. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel