On 2017/08/03 13:18, trondd wrote: > On Thu, August 3, 2017 1:09 pm, Stuart Henderson wrote: > > On 2017/08/03 09:55, trondd wrote: > >> I noticed, when applying multiple patches, I would eventually be told I > >> was out of space on /. There was about 60M free which is plenty for a > >> kernel and a copy. Rerunning syspatch would allow it to continue for a > >> another handful of patches. > >> > >> Mucking about in checkfs(), it seems that the value of the space needed > >> for a partition is not cleared and accumulates for each patch. Afer a > >> few > >> kernel patches, the number grows 10M at a time and I'd eventually be > >> told > >> I didn't have enough space. > > > > This was fixed (22 July), you can copy the syspatch script from a -current > > system to a 6.1. With 6.1's syspatch, unless you're very tight on space, > > you can keep re-running it and it should eventually make it through. > > > > I believe the clean ups were for data on disk. This problem seems to be > with the variable value in memory. My free disk space isn't changing, > just the free space syspatch thinks it needs.
Ah, perhaps the change to disk behaviour wasn't reflected in calculations then.. In any event, with 6.1 syspatch (which is the situation most people who now run into this will hit), you can just re-run it.
