Re: [libvirt PATCH v5 9/9] lxcDomainDetachDeviceHostdevUSBLive: Use VIR_WITH_OBJECT_LOCK_GUARD

On 2/2/22 01:06, Martin Kletzander wrote: > On Tue, Feb 01, 2022 at 05:23:33PM +0100, Tim Wiederhake wrote: >> On Tue, 2022-02-01 at 16:07 +0000, Daniel P. Berrangé wrote: >>> On Tue, Feb 01, 2022 at 03:25:55PM +0100, Martin Kletzander wrote: >>> > On Tue, Feb 01, 2022 at 02:20:17PM +0100, Tim Wiederhake wrote: >>> > > Signed-off-by: Tim Wiederhake <twiederh(a)redhat.com&gt; >>> > > --- >>> > > src/lxc/lxc_driver.c | 6 +++--- >>> > > 1 file changed, 3 insertions(+), 3 deletions(-) >>> > > >>> > > diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c >>> > > index 7bc39120ee..42053de9c3 100644 >>> > > --- a/src/lxc/lxc_driver.c >>> > > +++ b/src/lxc/lxc_driver.c >>> > > @@ -4045,9 +4045,9 @@ >>> > > lxcDomainDetachDeviceHostdevUSBLive(virLXCDriver *driver, >>> > >         VIR_WARN("cannot deny device %s for domain %s: %s", >>> > >                  dst, vm->def->name, virGetLastErrorMessage()); >>> > > >>> > > -    virObjectLock(hostdev_mgr->activeUSBHostdevs); >>> > > -    virUSBDeviceListDel(hostdev_mgr->activeUSBHostdevs, usb); >>> > > -    virObjectUnlock(hostdev_mgr->activeUSBHostdevs); >>> > > +    VIR_WITH_OBJECT_LOCK_GUARD(hostdev_mgr->activeUSBHostdevs) { >>> > > +        virUSBDeviceListDel(hostdev_mgr->activeUSBHostdevs, >>> > > usb); >>> > > +    } >>> > >>> > Even nicer you can omit the curly brackets and make this a +2/-3 ;) >>> > Otherwise I guess we'd need some kind of stylistic guide to keep us >>> > consistent in this regard right from the start. >>> >>> My preference is the keep the scope of the guarded section explicit >>> using {} even with a 1 line body. >> >> Same for me, fwiw. >> > > OK, I don't really have a horse in this race.  But in that case it would > be nice > to document it in our guidelines, maybe together with few words around > the new > macros.  In a later patch for example. Since you brought this up, I vaguely remember us discussing this earlier, though it was for one line body of if()-s. I too am in favor of having curly braces in that case, except maybe when the body is a goto or return statement. IOW: if (cond) return -1; // good if (cond) { func(); // good } if (cond) { goto cleanup; // not a big fan } OTOH, consistency matters more so I can live with the last example too.