Pass the drive's readonly attribute to the guest OS
PATCH 1/1 Pass the drive's readonly attribute to the guest OS
Hi,
as a preliminary step for adding read only flag for the -drive command, I've added code to pass the (existing) read only attribute of the drive to the guest OS (if it bother to ask).
I've added for virtio and for scsi.
Verified it already works for floppy.
Searched like mad (in linux sources) for same read only/write protect attribute for ide, but all I found was only for ide-floppy. I assume it's not supported in ide drives. Please correct me if I'm wrong.
I don't know about the other interface types for the drive command: sd, mtd, floppy, pflash. Where is usb ?? Will be happy to add those, too, if applicable. Please advise.
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
Naphtali
--- hw/scsi-disk.c | 3 ++- hw/virtio-blk.c | 3 +++ 2 files changed, 5 insertions(+), 1 deletions(-)
hw/scsi-disk.c +2 -1
@@ -631,7 +631,8 @@ static int32_t scsi_send_command(SCSIDevice *d, uint32_t tag,memset (p,0 ,4 ); outbuf[1 ] =0 ;/* Default media type. */ outbuf[3 ] =0 ;/* Block descriptor length. */ -if (bdrv_get_type_hint (s->dinfo->bdrv) ==BDRV_TYPE_CDROM ) { +if (bdrv_get_type_hint (s->dinfo->bdrv) ==BDRV_TYPE_CDROM || +bdrv_is_read_only (s->dinfo->bdrv)) { outbuf[2 ] =0x80 ;/* Readonly. */ } p +=4 ;
hw/virtio-blk.c +3 -0
@@ -444,6 +444,9 @@ static uint32_t virtio_blk_get_features(VirtIODevice *vdev)#endif if (strcmp (s->serial_str,"0" )) features |=1 <<VIRTIO_BLK_F_IDENTIFY ; + +if (bdrv_is_read_only (s->bs)) + features |=1 <<VIRTIO_BLK_F_RO ;return features; }
Signature · 1 line
1.6.3.3
Quoted 7 lines · Naphtali Sprei
Hi,
as a preliminary step for adding read only flag for the -drive command, I've added code to pass the (existing) read only attribute of the drive to the guest OS (if it bother to ask).I've added for virtio and for scsi.
Where is usb ??
hw/usb-msd.c
should be already covered by scsi though.
cheers,
Gerd
Quoted 7 lines · Gerd Hoffmann
Quoted 6 lines · Naphtali Sprei
Hi,
as a preliminary step for adding read only flag for the -drive command, I've added code to pass
the (existing) read only attribute of the drive to the guest OS (if it bother to ask).
Quoted 5 lines
Quoted 3 lines
I've added for virtio and for scsi.
Where is usb ??
hw/usb-msd.c
Thanks,
what I meant is I can't find how can it be specified as a drive interface.
Quoted 6 lines
should be already covered by scsi though.
cheers,
Gerd
Quoted 8 lines · Naphtali Sprei
Quoted 7 lines · Gerd Hoffmann
Quoted 6 lines · Naphtali Sprei
Hi,
as a preliminary step for adding read only flag for the -drive command, I've added code to pass
the (existing) read only attribute of the drive to the guest OS (if it bother to ask).
Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
I've added for virtio and for scsi.
Where is usb ??
hw/usb-msd.cThanks,
what I meant is I can't find how can it be specified as a drive interface.
You can use either
-usbdevice disk:/some/image
(usb_add in monitor)
or
-drive if=none,id=pendrive,file=/some/image
-device usb-storage,drive=pendrive
(drive_add + device_add in monitor)
cheers,
Gerd
Quoted 24 lines · Gerd Hoffmann
Quoted 10 lines · Naphtali Sprei
Quoted 9 lines · Gerd Hoffmann
Quoted 7 lines · Naphtali Sprei
Hi,
as a preliminary step for adding read only flag for the -drive command, I've added code to pass
the (existing) read only attribute of the drive to the guest OS (if it bother to ask).I've added for virtio and for scsi.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Where is usb ??
hw/usb-msd.cThanks,
what I meant is I can't find how can it be specified as a drive interface.You can use either
-usbdevice disk:/some/image
(usb_addin monitor)
Quoted 11 lines
or
-drive if=none,id=pendrive,file=/some/image
-device usb-storage,drive=pendrive
(drive_add+device_addin monitor)cheers,
Gerd
Thanks.
Verified, it "works", meaning readonly attribute passed to guest.
Naphtali Sprei wrote:
<snip>
Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
Naphtali
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 7 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
Kevin
Quoted 6 lines · Kevin Wolf
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 5 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
Quoted 7 lines
Quoted 2 lines
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
--
Gleb.
Am 15.10.2009 11:43, schrieb Gleb Natapov:
Quoted 10 lines · Kevin Wolf
Quoted 9 lines
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 7 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
Quoted 6 lines
Quoted 4 lines
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
Kevin
Quoted 15 lines · Kevin Wolf
Am 15.10.2009 11:43, schrieb Gleb Natapov:
Quoted 14 lines · Kevin Wolf
Quoted 13 lines
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 7 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
Quoted 7 lines
Quoted 2 lines
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
We pause only on host IO errors, but if we open underlying file as read only (do we?) and try to write into it we will get an IO error in the host.
--
Gleb.
Am 15.10.2009 11:54, schrieb Gleb Natapov:
Quoted 19 lines · Kevin Wolf
Quoted 18 lines
Am 15.10.2009 11:43, schrieb Gleb Natapov:
Quoted 16 lines · Kevin Wolf
Quoted 6 lines
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 5 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
Quoted 7 lines
Quoted 2 lines
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
Quoted 7 lines
Quoted 4 lines
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
We pause only on host IO errors, but if we open underlying file as read only (do we?) and try to write into it we will get an IO error in the host.
No, we'll return an error before a write request to the host is even issued.
Kevin
Quoted 24 lines · Kevin Wolf
Am 15.10.2009 11:54, schrieb Gleb Natapov:
Quoted 23 lines · Kevin Wolf
Quoted 22 lines
Am 15.10.2009 11:43, schrieb Gleb Natapov:
Quoted 10 lines · Kevin Wolf
Quoted 9 lines
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 7 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
Quoted 6 lines
Quoted 4 lines
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
Quoted 6 lines
Quoted 3 lines
We pause only on host IO errors, but if we open underlying file as read only (do we?) and try to write into it we will get an IO error in the host.
No, we'll return an error before a write request to the host is even issued.
Who is "we"? If "we" == "bdrv_write()/dma_bdrv_write()" then it's all the same. Upper layers don't actually care why block driver failed.
--
Gleb.
Am 15.10.2009 12:01, schrieb Gleb Natapov:
Quoted 25 lines · Kevin Wolf
Quoted 24 lines
Am 15.10.2009 11:54, schrieb Gleb Natapov:
Quoted 23 lines · Kevin Wolf
Quoted 22 lines
Am 15.10.2009 11:43, schrieb Gleb Natapov:
Quoted 10 lines · Kevin Wolf
Quoted 9 lines
Am 14.10.2009 18:40, schrieb Naphtali Sprei:
Quoted 7 lines
Naphtali Sprei wrote:
<snip>Quoted 2 lines
I'm planning to investigate where qemu should check the read only attribute before exeuting any write command to drives, would be sent in a different patch.
revisiting it, if guest OS knows it's a read only device and tries to modify it, anyhow, we don't really care about error reporting, as long as qemu doesn't crash (or modify the drive).
Quoted 6 lines
Quoted 4 lines
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
We pause only on host IO errors, but if we open underlying file as read only (do we?) and try to write into it we will get an IO error in the host.
No, we'll return an error before a write request to the host is even issued.
Who is "we"? If "we" == "bdrv_write()/dma_bdrv_write()" then it's all the same. Upper layers don't actually care why block driver failed.
Right, "we" is the qemu block layer. If the devices don't use the error code returned, they better should be fixed, I think?
Kevin
Quoted 15 lines · Kevin Wolf
Quoted 15 lines
Quoted 15 lines
Quoted 14 lines
Quoted 11 lines
Quoted 6 lines
Quoted 4 lines
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
We pause only on host IO errors, but if we open underlying file as read only (do we?) and try to write into it we will get an IO error in the host.
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
No, we'll return an error before a write request to the host is even issued.
Who is "we"? If "we" == "bdrv_write()/dma_bdrv_write()" then it's all the same. Upper layers don't actually care why block driver failed.
Right, "we" is the qemu block layer. If the devices don't use the error code returned, they better should be fixed, I think?
Fixed in what way? There is an option to pause VM on any write error. If block layer returns write error for whatever reason VM will be paused. If scsi/ide/virtio knows that the media is read only it shouldn't issue writes to block layer, but handle it like real HW would.
--
Gleb.
Am 15.10.2009 12:11, schrieb Gleb Natapov:
Quoted 21 lines · Kevin Wolf
Quoted 20 lines
Quoted 19 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 6 lines
Quoted 4 lines
If the right response to a write on a read-only device is defined in the specification (and it most probably is), we should still give the right response, even though the OS is doing something wrong.
And since our response to write error may be pausing a VM we shouldn't allow this to be triggered by a guest OS.
I thought we only pause the VM if we get an host IO error? But if you do want to stop it for all errors, you shouldn't start suppressing errors so that it doesn't stop.
Quoted 6 lines
Quoted 3 lines
We pause only on host IO errors, but if we open underlying file as read only (do we?) and try to write into it we will get an IO error in the host.
No, we'll return an error before a write request to the host is even issued.
Who is "we"? If "we" == "bdrv_write()/dma_bdrv_write()" then it's all the same. Upper layers don't actually care why block driver failed.
Quoted 7 lines
Quoted 3 lines
Right, "we" is the qemu block layer. If the devices don't use the error code returned, they better should be fixed, I think?
Fixed in what way? There is an option to pause VM on any write error. If block layer returns write error for whatever reason VM will be paused. If scsi/ide/virtio knows that the media is read only it shouldn't issue writes to block layer, but handle it like real HW would.
Ok, I agree if you want to do it this way. How it's done I really don't care too much, but in the end the guest OS should see the right error. I had the impression that Naphtali and you wanted to silently ignore writes to a read-only disk, which would be just wrong.
Kevin