| Message ID | 20220831165824.94815-2-sj@kernel.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | xen-blk{front,back}: Fix the broken semantic and flow of feature-persistent | expand |
On 31/08/22 04:58PM, SeongJae Park wrote: > The advertisement of the persistent grants feature (writing > 'feature-persistent' to xenbus) should mean not the decision for using > the feature but only the availability of the feature. However, commit > aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent > grants") made a field of blkback, which was a place for saving only the > negotiation result, to be used for yet another purpose: caching of the > 'feature_persistent' parameter value. As a result, the advertisement, > which should follow only the parameter value, becomes inconsistent. > > This commit fixes the misuse of the semantic by making blkback saves the > parameter value in a separate place and advertises the support based on > only the saved value. > > Fixes: aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent grants") > Cc: <stable@vger.kernel.org> # 5.10.x > Suggested-by: Juergen Gross <jgross@suse.com> > Signed-off-by: SeongJae Park <sj@kernel.org> > --- > drivers/block/xen-blkback/common.h | 3 +++ > drivers/block/xen-blkback/xenbus.c | 6 ++++-- > 2 files changed, 7 insertions(+), 2 deletions(-) > > diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h > index bda5c815e441..a28473470e66 100644 > --- a/drivers/block/xen-blkback/common.h > +++ b/drivers/block/xen-blkback/common.h > @@ -226,6 +226,9 @@ struct xen_vbd { > sector_t size; > unsigned int flush_support:1; > unsigned int discard_secure:1; > + /* Connect-time cached feature_persistent parameter value */ > + unsigned int feature_gnt_persistent_parm:1; Continuing over from the previous version: > > If feature_gnt_persistent_parm is always going to be equal to > > feature_persistent, then why introduce it at all? Why not just use > > feature_persistent directly? This way you avoid adding an extra flag > > whose purpose is not immediately clear, and you also avoid all the > > mess with setting this flag at the right time. > > Mainly because the parameter should read twice (once for > advertisement, and once later just before the negotitation, for > checking if we advertised or not), and the user might change the > parameter value between the two reads. > > For the detailed available sequence of the race, you could refer to the > prior conversation[1]. > > [1] https://lore.kernel.org/linux-block/20200922111259.GJ19254@Air-de-Roger/ Okay, I see. Thanks for the pointer. But still, I think it would be better to not maintain two copies of the value. How about doing: blkif->vbd.feature_gnt_persistent = xenbus_read_unsigned(dev->nodename, "feature-persistent", 0) && xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); This makes it quite clear that we only enable persistent grants if _both_ ends support it. We can do the same for blkfront. > + /* Persistent grants feature negotiation result */ > unsigned int feature_gnt_persistent:1; > unsigned int overflow_max_grants:1; > }; > diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c > index ee7ad2fb432d..c0227dfa4688 100644 > --- a/drivers/block/xen-blkback/xenbus.c > +++ b/drivers/block/xen-blkback/xenbus.c > @@ -907,7 +907,7 @@ static void connect(struct backend_info *be) > xen_blkbk_barrier(xbt, be, be->blkif->vbd.flush_support); > > err = xenbus_printf(xbt, dev->nodename, "feature-persistent", "%u", > - be->blkif->vbd.feature_gnt_persistent); > + be->blkif->vbd.feature_gnt_persistent_parm); > if (err) { > xenbus_dev_fatal(dev, err, "writing %s/feature-persistent", > dev->nodename); > @@ -1085,7 +1085,9 @@ static int connect_ring(struct backend_info *be) > return -ENOSYS; > } > > - blkif->vbd.feature_gnt_persistent = feature_persistent && > + blkif->vbd.feature_gnt_persistent_parm = feature_persistent; > + blkif->vbd.feature_gnt_persistent = > + blkif->vbd.feature_gnt_persistent_parm && > xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); > > blkif->vbd.overflow_max_grants = 0; > -- > 2.25.1 >
On 02.09.22 11:53, Pratyush Yadav wrote: > On 31/08/22 04:58PM, SeongJae Park wrote: >> The advertisement of the persistent grants feature (writing >> 'feature-persistent' to xenbus) should mean not the decision for using >> the feature but only the availability of the feature. However, commit >> aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent >> grants") made a field of blkback, which was a place for saving only the >> negotiation result, to be used for yet another purpose: caching of the >> 'feature_persistent' parameter value. As a result, the advertisement, >> which should follow only the parameter value, becomes inconsistent. >> >> This commit fixes the misuse of the semantic by making blkback saves the >> parameter value in a separate place and advertises the support based on >> only the saved value. >> >> Fixes: aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent grants") >> Cc: <stable@vger.kernel.org> # 5.10.x >> Suggested-by: Juergen Gross <jgross@suse.com> >> Signed-off-by: SeongJae Park <sj@kernel.org> >> --- >> drivers/block/xen-blkback/common.h | 3 +++ >> drivers/block/xen-blkback/xenbus.c | 6 ++++-- >> 2 files changed, 7 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h >> index bda5c815e441..a28473470e66 100644 >> --- a/drivers/block/xen-blkback/common.h >> +++ b/drivers/block/xen-blkback/common.h >> @@ -226,6 +226,9 @@ struct xen_vbd { >> sector_t size; >> unsigned int flush_support:1; >> unsigned int discard_secure:1; >> + /* Connect-time cached feature_persistent parameter value */ >> + unsigned int feature_gnt_persistent_parm:1; > > Continuing over from the previous version: > >>> If feature_gnt_persistent_parm is always going to be equal to >>> feature_persistent, then why introduce it at all? Why not just use >>> feature_persistent directly? This way you avoid adding an extra flag >>> whose purpose is not immediately clear, and you also avoid all the >>> mess with setting this flag at the right time. >> >> Mainly because the parameter should read twice (once for >> advertisement, and once later just before the negotitation, for >> checking if we advertised or not), and the user might change the >> parameter value between the two reads. >> >> For the detailed available sequence of the race, you could refer to the >> prior conversation[1]. >> >> [1] https://lore.kernel.org/linux-block/20200922111259.GJ19254@Air-de-Roger/ > > Okay, I see. Thanks for the pointer. But still, I think it would be > better to not maintain two copies of the value. How about doing: > > blkif->vbd.feature_gnt_persistent = > xenbus_read_unsigned(dev->nodename, "feature-persistent", 0) && > xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); > > This makes it quite clear that we only enable persistent grants if > _both_ ends support it. We can do the same for blkfront. I prefer it as is, as it will not rely on nobody having modified the Xenstore node (which would in theory be possible). Juergen
On Fri, Sep 02, 2022 at 09:53:22AM +0000, Pratyush Yadav wrote: > On 31/08/22 04:58PM, SeongJae Park wrote: > > The advertisement of the persistent grants feature (writing > > 'feature-persistent' to xenbus) should mean not the decision for using > > the feature but only the availability of the feature. However, commit > > aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent > > grants") made a field of blkback, which was a place for saving only the > > negotiation result, to be used for yet another purpose: caching of the > > 'feature_persistent' parameter value. As a result, the advertisement, > > which should follow only the parameter value, becomes inconsistent. > > > > This commit fixes the misuse of the semantic by making blkback saves the > > parameter value in a separate place and advertises the support based on > > only the saved value. > > > > Fixes: aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent grants") > > Cc: <stable@vger.kernel.org> # 5.10.x > > Suggested-by: Juergen Gross <jgross@suse.com> > > Signed-off-by: SeongJae Park <sj@kernel.org> > > --- > > drivers/block/xen-blkback/common.h | 3 +++ > > drivers/block/xen-blkback/xenbus.c | 6 ++++-- > > 2 files changed, 7 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h > > index bda5c815e441..a28473470e66 100644 > > --- a/drivers/block/xen-blkback/common.h > > +++ b/drivers/block/xen-blkback/common.h > > @@ -226,6 +226,9 @@ struct xen_vbd { > > sector_t size; > > unsigned int flush_support:1; > > unsigned int discard_secure:1; > > + /* Connect-time cached feature_persistent parameter value */ > > + unsigned int feature_gnt_persistent_parm:1; > > Continuing over from the previous version: > > > > If feature_gnt_persistent_parm is always going to be equal to > > > feature_persistent, then why introduce it at all? Why not just use > > > feature_persistent directly? This way you avoid adding an extra flag > > > whose purpose is not immediately clear, and you also avoid all the > > > mess with setting this flag at the right time. > > > > Mainly because the parameter should read twice (once for > > advertisement, and once later just before the negotitation, for > > checking if we advertised or not), and the user might change the > > parameter value between the two reads. > > > > For the detailed available sequence of the race, you could refer to the > > prior conversation[1]. > > > > [1] https://lore.kernel.org/linux-block/20200922111259.GJ19254@Air-de-Roger/ > > Okay, I see. Thanks for the pointer. But still, I think it would be > better to not maintain two copies of the value. How about doing: > > blkif->vbd.feature_gnt_persistent = > xenbus_read_unsigned(dev->nodename, "feature-persistent", 0) && > xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); > > This makes it quite clear that we only enable persistent grants if > _both_ ends support it. We can do the same for blkfront. Currently, the feature-persistent xenstore entry is written to from connect() which is called after connect_ring(). So it's not available like this. Perhaps it's possible to delay the decision whether to use persistent grants until connect(). > > > + /* Persistent grants feature negotiation result */ > > unsigned int feature_gnt_persistent:1; > > unsigned int overflow_max_grants:1; > > }; > > diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c > > index ee7ad2fb432d..c0227dfa4688 100644 > > --- a/drivers/block/xen-blkback/xenbus.c > > +++ b/drivers/block/xen-blkback/xenbus.c > > @@ -907,7 +907,7 @@ static void connect(struct backend_info *be) > > xen_blkbk_barrier(xbt, be, be->blkif->vbd.flush_support); > > > > err = xenbus_printf(xbt, dev->nodename, "feature-persistent", "%u", > > - be->blkif->vbd.feature_gnt_persistent); > > + be->blkif->vbd.feature_gnt_persistent_parm); > > if (err) { > > xenbus_dev_fatal(dev, err, "writing %s/feature-persistent", > > dev->nodename); > > @@ -1085,7 +1085,9 @@ static int connect_ring(struct backend_info *be) > > return -ENOSYS; > > } > > > > - blkif->vbd.feature_gnt_persistent = feature_persistent && > > + blkif->vbd.feature_gnt_persistent_parm = feature_persistent; > > + blkif->vbd.feature_gnt_persistent = > > + blkif->vbd.feature_gnt_persistent_parm && > > xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); > > > > blkif->vbd.overflow_max_grants = 0; > > -- > > 2.25.1 > > > > -- > Amazon Development Center Germany GmbH > Krausenstr. 38 > 10117 Berlin > Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss > Eingetragen am Amtsgericht Charlottenburg unter HRB 149173 B > Sitz: Berlin > Ust-ID: DE 289 237 879 Amazon Development Center Germany GmbH Krausenstr. 38 10117 Berlin Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss Eingetragen am Amtsgericht Charlottenburg unter HRB 149173 B Sitz: Berlin Ust-ID: DE 289 237 879
On 02/09/22 01:08PM, Juergen Gross wrote: > On 02.09.22 11:53, Pratyush Yadav wrote: > > On 31/08/22 04:58PM, SeongJae Park wrote: > > > The advertisement of the persistent grants feature (writing > > > 'feature-persistent' to xenbus) should mean not the decision for using > > > the feature but only the availability of the feature. However, commit > > > aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent > > > grants") made a field of blkback, which was a place for saving only the > > > negotiation result, to be used for yet another purpose: caching of the > > > 'feature_persistent' parameter value. As a result, the advertisement, > > > which should follow only the parameter value, becomes inconsistent. > > > > > > This commit fixes the misuse of the semantic by making blkback saves the > > > parameter value in a separate place and advertises the support based on > > > only the saved value. > > > > > > Fixes: aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent grants") > > > Cc: <stable@vger.kernel.org> # 5.10.x > > > Suggested-by: Juergen Gross <jgross@suse.com> > > > Signed-off-by: SeongJae Park <sj@kernel.org> > > > --- > > > drivers/block/xen-blkback/common.h | 3 +++ > > > drivers/block/xen-blkback/xenbus.c | 6 ++++-- > > > 2 files changed, 7 insertions(+), 2 deletions(-) > > > > > > diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h > > > index bda5c815e441..a28473470e66 100644 > > > --- a/drivers/block/xen-blkback/common.h > > > +++ b/drivers/block/xen-blkback/common.h > > > @@ -226,6 +226,9 @@ struct xen_vbd { > > > sector_t size; > > > unsigned int flush_support:1; > > > unsigned int discard_secure:1; > > > + /* Connect-time cached feature_persistent parameter value */ > > > + unsigned int feature_gnt_persistent_parm:1; > > > > Continuing over from the previous version: > > > > > > If feature_gnt_persistent_parm is always going to be equal to > > > > feature_persistent, then why introduce it at all? Why not just use > > > > feature_persistent directly? This way you avoid adding an extra flag > > > > whose purpose is not immediately clear, and you also avoid all the > > > > mess with setting this flag at the right time. > > > > > > Mainly because the parameter should read twice (once for > > > advertisement, and once later just before the negotitation, for > > > checking if we advertised or not), and the user might change the > > > parameter value between the two reads. > > > > > > For the detailed available sequence of the race, you could refer to the > > > prior conversation[1]. > > > > > > [1] https://lore.kernel.org/linux-block/20200922111259.GJ19254@Air-de-Roger/ > > > > Okay, I see. Thanks for the pointer. But still, I think it would be > > better to not maintain two copies of the value. How about doing: > > > > blkif->vbd.feature_gnt_persistent = > > xenbus_read_unsigned(dev->nodename, "feature-persistent", 0) && > > xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); > > > > This makes it quite clear that we only enable persistent grants if > > _both_ ends support it. We can do the same for blkfront. > > I prefer it as is, as it will not rely on nobody having modified the > Xenstore node (which would in theory be possible). Okay. In that case, Reviewed-by: Pratyush Yadav <ptyadav@amazon.de>
diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h index bda5c815e441..a28473470e66 100644 --- a/drivers/block/xen-blkback/common.h +++ b/drivers/block/xen-blkback/common.h @@ -226,6 +226,9 @@ struct xen_vbd { sector_t size; unsigned int flush_support:1; unsigned int discard_secure:1; + /* Connect-time cached feature_persistent parameter value */ + unsigned int feature_gnt_persistent_parm:1; + /* Persistent grants feature negotiation result */ unsigned int feature_gnt_persistent:1; unsigned int overflow_max_grants:1; }; diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c index ee7ad2fb432d..c0227dfa4688 100644 --- a/drivers/block/xen-blkback/xenbus.c +++ b/drivers/block/xen-blkback/xenbus.c @@ -907,7 +907,7 @@ static void connect(struct backend_info *be) xen_blkbk_barrier(xbt, be, be->blkif->vbd.flush_support); err = xenbus_printf(xbt, dev->nodename, "feature-persistent", "%u", - be->blkif->vbd.feature_gnt_persistent); + be->blkif->vbd.feature_gnt_persistent_parm); if (err) { xenbus_dev_fatal(dev, err, "writing %s/feature-persistent", dev->nodename); @@ -1085,7 +1085,9 @@ static int connect_ring(struct backend_info *be) return -ENOSYS; } - blkif->vbd.feature_gnt_persistent = feature_persistent && + blkif->vbd.feature_gnt_persistent_parm = feature_persistent; + blkif->vbd.feature_gnt_persistent = + blkif->vbd.feature_gnt_persistent_parm && xenbus_read_unsigned(dev->otherend, "feature-persistent", 0); blkif->vbd.overflow_max_grants = 0;
The advertisement of the persistent grants feature (writing 'feature-persistent' to xenbus) should mean not the decision for using the feature but only the availability of the feature. However, commit aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent grants") made a field of blkback, which was a place for saving only the negotiation result, to be used for yet another purpose: caching of the 'feature_persistent' parameter value. As a result, the advertisement, which should follow only the parameter value, becomes inconsistent. This commit fixes the misuse of the semantic by making blkback saves the parameter value in a separate place and advertises the support based on only the saved value. Fixes: aac8a70db24b ("xen-blkback: add a parameter for disabling of persistent grants") Cc: <stable@vger.kernel.org> # 5.10.x Suggested-by: Juergen Gross <jgross@suse.com> Signed-off-by: SeongJae Park <sj@kernel.org> --- drivers/block/xen-blkback/common.h | 3 +++ drivers/block/xen-blkback/xenbus.c | 6 ++++-- 2 files changed, 7 insertions(+), 2 deletions(-)