All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ken <jinjian@huawei.com>
To: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: <xen-devel@lists.xenproject.org>, <axboe@suse.de>,
	<rth@twiddle.net>, <linux-kernel@vger.kernel.org>,
	<xuquan8@huawei.com>, <zhang.zhanghailiang@huawei.com>,
	<pss.wulizhen@huawei.com>
Subject: Re: [Xen-devel] Xen/cdrom: Ubuntu 16.04 VM read the content from CD-ROM abnormally
Date: Mon, 12 Dec 2016 09:22:27 +0800	[thread overview]
Message-ID: <584DFBD3.6090406@huawei.com> (raw)
In-Reply-To: <20161209161543.GB11877@char.us.oracle.com>



On 2016/12/10 0:15, Konrad Rzeszutek Wilk wrote:
> On Fri, Dec 09, 2016 at 04:21:02PM +0800, Ken wrote:
>> Hi all,
>>
>> I run the Ubuntu 16.04 server (2 vcpu/2G, Linux 4.4.0) on the Xen-4.1.2, and
>> installed gcc through the CDROM used by 16.04 iso file, when I installed gcc
>> that depends deb packages to decompress failed. But uploaded the ISO files
>> into the VM are mounted by loop or used as CDROM for other VM (Fedora 24)
>> transport these abnormal files to 16.04 are available. So the 16.04 ISO file
>> should be correct.
>>
>> Then I went to try to fix this problem, the steps are as follows:
>> First of all, I was worried because the Hypervisor version is too old to
>> cause this problems, so I upgraded to the Xen upstream and found that there
>> are the problem still.
>>
>> Then, I went to attempted to upgrade the VM kernel, by dichotomy to find the
>> smallest available kernel, but found that the kernel patch has nothing to do
>> the cdrom or isofs driver, Linux kernel committed history has nothing to do
>> these, combined with the Fedora within Linux 4.4.0 have no problem, so I
>> inferred the kernel had no this defective.
>>
>> The above comparison can not troubleshoot the problem, so I analyzed the
>> Linux CD-ROM device loading process to confirm that the problem is
>> encountered before the mount. and I found the following three strange
>> phenomenon:
>> 1. Using Ubuntu 14.04.5 udevadm, re-packaged the init-ramfs of 16.04, reboot
>> the VM, read the content of CD-ROM successfully, but compile the system-204
>> from the 14.04.5 and installed on 16.04, the problem still can not be
>> resolved, some abnormal logs as blow:
>>     random: udevadm: uninitialized urandom read (16 bytes read, 28 bits of
>> entropy available)
>>
>> 2. found that the VM registers dmi failed at startup, so I removed the dmi
>> driver, reboot the VM, read the content of CD-ROM successfully, and from the
>> kernel dmi_scan_machine code, it should not affect to use the CD-ROM, The
>> failed logs as blow:
>>     ioremap error for 0xfc001000-0xfc002000, requested 0x2, got 0x0
>>     dmi: Firmware registration failed.
>>
>> 3. Deploy the VM within 8 vcpus, 8G memory, read the content of CD-ROM
>> successfully.
>>
>> Therefore, I inferred this anomaly that has several problems, but can not
>> focus it, there are other people have encountered this same anomaly? Do
>> anyone have any suggests to
>> debug it next step?
>
> I think you are the first one to see this. I had run small guests before
> (1GB, 1VCPU) with ISO and did not encounter these issues. Are there
> erorrs in the dmesg when the file is copied?

The other VM are normal, it is only with Ubuntu 16.04 and its 
LTS(smaller than 8GB, 8VCPU). sometimes, the ubuntu 16.04 encountered 
problem that maybe printed some logs as blow:
    ISOFS: Interleaved files not (yet) supported.
    ISOFS: inode->i_ino = (14976).
    ISOFS: File unit size != 0 for ISO file (14976).

>>
>> Thanks.
>>
>>
>>
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xen.org
>> https://lists.xen.org/xen-devel
>
> .
>

WARNING: multiple messages have this Message-ID (diff)
From: Ken <jinjian@huawei.com>
To: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: xuquan8@huawei.com, zhang.zhanghailiang@huawei.com,
	linux-kernel@vger.kernel.org, pss.wulizhen@huawei.com,
	xen-devel@lists.xenproject.org, axboe@suse.de, rth@twiddle.net
Subject: Re: Xen/cdrom: Ubuntu 16.04 VM read the content from CD-ROM abnormally
Date: Mon, 12 Dec 2016 09:22:27 +0800	[thread overview]
Message-ID: <584DFBD3.6090406@huawei.com> (raw)
In-Reply-To: <20161209161543.GB11877@char.us.oracle.com>



On 2016/12/10 0:15, Konrad Rzeszutek Wilk wrote:
> On Fri, Dec 09, 2016 at 04:21:02PM +0800, Ken wrote:
>> Hi all,
>>
>> I run the Ubuntu 16.04 server (2 vcpu/2G, Linux 4.4.0) on the Xen-4.1.2, and
>> installed gcc through the CDROM used by 16.04 iso file, when I installed gcc
>> that depends deb packages to decompress failed. But uploaded the ISO files
>> into the VM are mounted by loop or used as CDROM for other VM (Fedora 24)
>> transport these abnormal files to 16.04 are available. So the 16.04 ISO file
>> should be correct.
>>
>> Then I went to try to fix this problem, the steps are as follows:
>> First of all, I was worried because the Hypervisor version is too old to
>> cause this problems, so I upgraded to the Xen upstream and found that there
>> are the problem still.
>>
>> Then, I went to attempted to upgrade the VM kernel, by dichotomy to find the
>> smallest available kernel, but found that the kernel patch has nothing to do
>> the cdrom or isofs driver, Linux kernel committed history has nothing to do
>> these, combined with the Fedora within Linux 4.4.0 have no problem, so I
>> inferred the kernel had no this defective.
>>
>> The above comparison can not troubleshoot the problem, so I analyzed the
>> Linux CD-ROM device loading process to confirm that the problem is
>> encountered before the mount. and I found the following three strange
>> phenomenon:
>> 1. Using Ubuntu 14.04.5 udevadm, re-packaged the init-ramfs of 16.04, reboot
>> the VM, read the content of CD-ROM successfully, but compile the system-204
>> from the 14.04.5 and installed on 16.04, the problem still can not be
>> resolved, some abnormal logs as blow:
>>     random: udevadm: uninitialized urandom read (16 bytes read, 28 bits of
>> entropy available)
>>
>> 2. found that the VM registers dmi failed at startup, so I removed the dmi
>> driver, reboot the VM, read the content of CD-ROM successfully, and from the
>> kernel dmi_scan_machine code, it should not affect to use the CD-ROM, The
>> failed logs as blow:
>>     ioremap error for 0xfc001000-0xfc002000, requested 0x2, got 0x0
>>     dmi: Firmware registration failed.
>>
>> 3. Deploy the VM within 8 vcpus, 8G memory, read the content of CD-ROM
>> successfully.
>>
>> Therefore, I inferred this anomaly that has several problems, but can not
>> focus it, there are other people have encountered this same anomaly? Do
>> anyone have any suggests to
>> debug it next step?
>
> I think you are the first one to see this. I had run small guests before
> (1GB, 1VCPU) with ISO and did not encounter these issues. Are there
> erorrs in the dmesg when the file is copied?

The other VM are normal, it is only with Ubuntu 16.04 and its 
LTS(smaller than 8GB, 8VCPU). sometimes, the ubuntu 16.04 encountered 
problem that maybe printed some logs as blow:
    ISOFS: Interleaved files not (yet) supported.
    ISOFS: inode->i_ino = (14976).
    ISOFS: File unit size != 0 for ISO file (14976).

>>
>> Thanks.
>>
>>
>>
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xen.org
>> https://lists.xen.org/xen-devel
>
> .
>


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

  reply	other threads:[~2016-12-12  1:23 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-12-09  8:21 Xen/cdrom: Ubuntu 16.04 VM read the content from CD-ROM abnormally Ken
2016-12-09 16:15 ` Konrad Rzeszutek Wilk
2016-12-09 16:15 ` [Xen-devel] " Konrad Rzeszutek Wilk
2016-12-12  1:22   ` Ken [this message]
2016-12-12  1:22     ` Ken

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=584DFBD3.6090406@huawei.com \
    --to=jinjian@huawei.com \
    --cc=axboe@suse.de \
    --cc=konrad.wilk@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pss.wulizhen@huawei.com \
    --cc=rth@twiddle.net \
    --cc=xen-devel@lists.xenproject.org \
    --cc=xuquan8@huawei.com \
    --cc=zhang.zhanghailiang@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.