On Apr 1, 2011, at 8:28 PM, NormW wrote: > G/M... > On 2/04/2011 5:26 AM, Jim Jagielski wrote: >> The consumers in the patch depend on the shm version of the >> slotmem impl... no idea what the impact of moving to plain >> would be, but functionality will be lost... > without: >> + if (!storage) >> + storage = ap_lookup_provider(AP_SLOTMEM_PROVIDER_GROUP, "plain", "0"); > > "plain" won't even be found, which is really lost.
My point is those consumers in the patch were designed to use shm; I doubt if they would work correctly if at all using the plain slotmem providers. If they *do* then I'm +1 on the patch, but adding the patch w/o verification is just silly...
