[Qemu-devel] commit 0be63901d2a33a6ed25caa5df3f530df75338f6a breaks "-kernel"

[Qemu-devel] commit 0be63901d2a33a6ed25caa5df3f530df75338f6a breaks "-kernel"

[Gabriel L. Somlo]

Using "-kenel <file>" results in a segmentation fault (at least with
qemu-system-x86_64). Bisect says the breakage was introduced by commit

commit 0be63901d2a33a6ed25caa5df3f530df75338f6a
Author: Gonglei <arei.gonglei@huawei.com>
Date:   Thu Jan 29 15:08:51 2015 +0800

    qdev: support to get a device firmware path directly
    
    commit 6b1566c (qdev: Introduce FWPathProvider interface) did a
    good job for supproting to get firmware path on some different
    architectures.
    
    Moreover further more, we can use the interface to get firmware
    path name for a device which isn't attached a specific bus,
    such as virtio-bus, scsi-bus etc.
    
    When the device (such as vhost-scsi) realize the TYPE_FW_PATH_PROVIDER
    interface, we should introduce a new function to get the correct firmware
    path name for it.
    
    Signed-off-by: Gonglei <arei.gonglei@huawei.com>
    Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Reverting the commit seems to fix the problem for me, as a temporary
solution.

Thanks,
--Gabriel

Re: [Qemu-devel] commit 0be63901d2a33a6ed25caa5df3f530df75338f6a breaks "-kernel"

[Paolo Bonzini]

On 02/03/2015 15:41, Gabriel L. Somlo wrote:
> Using "-kenel <file>" results in a segmentation fault (at least with
> qemu-system-x86_64). Bisect says the breakage was introduced by commit
> 
> commit 0be63901d2a33a6ed25caa5df3f530df75338f6a
> Author: Gonglei <arei.gonglei@huawei.com>
> Date:   Thu Jan 29 15:08:51 2015 +0800
> 
>     qdev: support to get a device firmware path directly
>     
>     commit 6b1566c (qdev: Introduce FWPathProvider interface) did a
>     good job for supproting to get firmware path on some different
>     architectures.
>     
>     Moreover further more, we can use the interface to get firmware
>     path name for a device which isn't attached a specific bus,
>     such as virtio-bus, scsi-bus etc.
>     
>     When the device (such as vhost-scsi) realize the TYPE_FW_PATH_PROVIDER
>     interface, we should introduce a new function to get the correct firmware
>     path name for it.
>     
>     Signed-off-by: Gonglei <arei.gonglei@huawei.com>
>     Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> 
> Reverting the commit seems to fix the problem for me, as a temporary
> solution.

Gonglei sent a patch for this, and it's in my pull request from this
morning.

Paolo