How does one add multiple bugs to the same CR?

papertigers: Just put one bug/line in the commit message.

rmustacc: thanks, wasn't sure if that was correct for how do commit messages in gate

Once a cr's up we can always change it, so don't worry too much.

Hmm I winder if Ian on the list is also bitten by the same viona issue i always am, forgot where I got stuck with illumos.org/issues/15464 ...

→ BUG 15464: Change the viona default to copy tx buffers instead of loaning them by default (New)

I wonder were my dev VM went too

I haven't found the VM but I did remember where I got stuck on, I was on sure how to deal with the logic in viona_tx_copy_needed() I could just add a new flag 'always copy' that defaults to try as propose in the original diff in illumos.org/issues/15271

→ FEATURE 15271: Want tunable to force viona to always copy tx buffers (New) | code.illumos.org/c/illumos-gate/+/2565

That also allows an easy out that would still do the proper thing on bnxe

But not sure that would be good enough

As that was originally rejected in favor of just documenting how to set the other enum.

[illumos-gate] 15815 Errors in the urtw(4d) manual -- Peter Tribble <peter.tribble⊙gc>

dos bhyve have gdb stub?

does*

ah, ok -G.

papertigers: we can even leave comments on your commit message as part of code review, no need to worry about it in gerrit. :)

richlowe: thanks, that's how I got myself duplicate CRs. Or I pushed to the wrong ref

the changeid needs to be constant, for updates.

yeah I should have the hook for it from previous gerrit usage.

hm: ERROR:root:Error b'bhyve: gdb address resolution: non-recoverable name resolution failure

I doubt it's tested much

seems so:(

though I desparately wish there was an mdb backend to talk to a gdb stub

which I bet would perk up the old timers.

there is very sketchy patch from jeffpc;)

never had time to get it to do useful work:D

I've used the qemu gdb stub quite a bit, but never with an accelerated backend like kvm.

yes, one in qemu is quite nice.

one in vmware fusion is quite buggy

I would like to ask some question about oxide’s crucible project, is there anyone could answer?

I’m going to test crucible and use it as my backup / archive / database space cross physical node, but I don’t know how to use it; I have 3 postgresql node(smartos vm), each postgresql node had it’s space for database, the same archive/backup space to store archive log and database backup; if then, how many downstairs/upstairs instance will be create? how to link the space to database vm? best wishes