block/ssh.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++++------ 1 file changed, 65 insertions(+), 7 deletions(-)
This series implements .bdrv_refresh_filename() for the ssh block driver, along with an appropriate .bdrv_dirname() so we don't chop off query strings for backing files with relative filenames. This series depends on my “block: Fix some filename generation issues” series and on Pino's “ssh: switch from libssh2 to libssh” patch. Based-on: 20180205151835.20812-1-mreitz@redhat.com Based-on: 20180118164439.2120-1-ptoscano@redhat.com Max Reitz (2): block/ssh: Implement .bdrv_refresh_filename() block/ssh: Implement .bdrv_dirname() block/ssh.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++++------ 1 file changed, 65 insertions(+), 7 deletions(-) -- 2.14.3
On Mon, Feb 05, 2018 at 09:22:30PM +0100, Max Reitz wrote: > This series implements .bdrv_refresh_filename() for the ssh block > driver, along with an appropriate .bdrv_dirname() so we don't chop off > query strings for backing files with relative filenames. > > This series depends on my “block: Fix some filename generation issues” > series and on Pino's “ssh: switch from libssh2 to libssh” patch. > > Based-on: 20180205151835.20812-1-mreitz@redhat.com > Based-on: 20180118164439.2120-1-ptoscano@redhat.com > > > Max Reitz (2): > block/ssh: Implement .bdrv_refresh_filename() > block/ssh: Implement .bdrv_dirname() > > block/ssh.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++++------ I reviewed the patches and I can't see anything wrong with them. I guess that on "weird" remote machines getting the directory from a filename is an impossible mission, but for reasonable hosts it's going to be fine. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
I assume the context of all this is making relative backing file names work in qcow2 files? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-builder quickly builds VMs from scratch http://libguestfs.org/virt-builder.1.html
On 2018-02-05 21:46, Richard W.M. Jones wrote:
>
> I assume the context of all this is making relative backing file names work
> in qcow2 files?
Yes, mostly. (https://bugzilla.redhat.com/show_bug.cgi?id=1528931)
But it's generally nice for block nodes to print a readable filename
when queried and not a json:{} monster all of the time.
Max
On 02/05/2018 03:22 PM, Max Reitz wrote: > This series implements .bdrv_refresh_filename() for the ssh block > driver, along with an appropriate .bdrv_dirname() so we don't chop off > query strings for backing files with relative filenames. > > This series depends on my “block: Fix some filename generation issues” > series and on Pino's “ssh: switch from libssh2 to libssh” patch. > > Based-on: 20180205151835.20812-1-mreitz@redhat.com > Based-on: 20180118164439.2120-1-ptoscano@redhat.com > > > Max Reitz (2): > block/ssh: Implement .bdrv_refresh_filename() > block/ssh: Implement .bdrv_dirname() > > block/ssh.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++++------ > 1 file changed, 65 insertions(+), 7 deletions(-) > Did this one rot on the vine? >1 month old.
Am 06.03.2018 um 22:51 hat John Snow geschrieben: > On 02/05/2018 03:22 PM, Max Reitz wrote: > > This series implements .bdrv_refresh_filename() for the ssh block > > driver, along with an appropriate .bdrv_dirname() so we don't chop off > > query strings for backing files with relative filenames. > > > > This series depends on my “block: Fix some filename generation issues” > > series and on Pino's “ssh: switch from libssh2 to libssh” patch. > > > > Based-on: 20180205151835.20812-1-mreitz@redhat.com > > Based-on: 20180118164439.2120-1-ptoscano@redhat.com > > > > > > Max Reitz (2): > > block/ssh: Implement .bdrv_refresh_filename() > > block/ssh: Implement .bdrv_dirname() > > > > block/ssh.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++++------ > > 1 file changed, 65 insertions(+), 7 deletions(-) > > Did this one rot on the vine? > > >1 month old. The Based-on tags are the problem, in particular the first one. But yes, we could possibly do more to review the dependencies... Kevin
On 03/07/2018 05:04 AM, Kevin Wolf wrote: > Am 06.03.2018 um 22:51 hat John Snow geschrieben: >> On 02/05/2018 03:22 PM, Max Reitz wrote: >>> This series implements .bdrv_refresh_filename() for the ssh block >>> driver, along with an appropriate .bdrv_dirname() so we don't chop off >>> query strings for backing files with relative filenames. >>> >>> This series depends on my “block: Fix some filename generation issues” >>> series and on Pino's “ssh: switch from libssh2 to libssh” patch. >>> >>> Based-on: 20180205151835.20812-1-mreitz@redhat.com >>> Based-on: 20180118164439.2120-1-ptoscano@redhat.com >>> >>> >>> Max Reitz (2): >>> block/ssh: Implement .bdrv_refresh_filename() >>> block/ssh: Implement .bdrv_dirname() >>> >>> block/ssh.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++++------ >>> 1 file changed, 65 insertions(+), 7 deletions(-) >> >> Did this one rot on the vine? >> >>> 1 month old. > > The Based-on tags are the problem, in particular the first one. But yes, > we could possibly do more to review the dependencies... > > Kevin > Yeah, sorry, I'm not pulling my review weight right now. I do try to keep my block-devel inbox a rough "todo," though, so for patches that get to that one month mark with no conclusion I like to ping them before I archive them. It's something I'd like to see patchew do, actually: "Here's a list of what's on the list that has no reviews or NACKs, and needs some love" coupled with a 30 day "Hey, nobody looked at this" ping to the list before it NACKs a set for being too old. I hope nobody reads these >1 month pings as me wondering why nobody ELSE has reviewed it, which is not my intent... --js
On Wed, 03/07 12:50, John Snow wrote: > It's something I'd like to see patchew do, actually: > > "Here's a list of what's on the list that has no reviews or NACKs, and > needs some love" It's not hard to define a search condition for that: http://patchew.org/search-help http://patchew.org/search?q=project%3AQEMU+age%3A%3E1m+not%3Areviewed+not%3Areplied+not%3Amerged+is%3Atested+to%3Aqemu-block > > coupled with a 30 day "Hey, nobody looked at this" ping to the list > before it NACKs a set for being too old. If the initial landing of the patch didn't get enough attention, chances are the pings will not change much about it especially it's from a bot. A summary list sounds good, though. Fam
On 03/07/2018 09:12 PM, Fam Zheng wrote: > On Wed, 03/07 12:50, John Snow wrote: >> It's something I'd like to see patchew do, actually: >> >> "Here's a list of what's on the list that has no reviews or NACKs, and >> needs some love" > > It's not hard to define a search condition for that: > > http://patchew.org/search-help > > http://patchew.org/search?q=project%3AQEMU+age%3A%3E1m+not%3Areviewed+not%3Areplied+not%3Amerged+is%3Atested+to%3Aqemu-block > >> >> coupled with a 30 day "Hey, nobody looked at this" ping to the list >> before it NACKs a set for being too old. > > If the initial landing of the patch didn't get enough attention, chances are the > pings will not change much about it especially it's from a bot. > > A summary list sounds good, though. > > Fam > It's not really to *get* the patch attention as much as it's meant to inform the sender that it's not under consideration anymore. "Yeah, but everyone knows this...?" Consider it a formalization of the process. --js
© 2016 - 2024 Red Hat, Inc.