Re: [Valgrind-users] Uniformity of inter-snapshot times in Massif

2009-09-01 Thread Belmon, Stephane
Added, bug # 205937. -- Stephane Belmon -Original Message- From: Nicholas Nethercote [mailto:n.netherc...@gmail.com] Sent: Tuesday, September 01, 2009 3:21 PM To: Belmon, Stephane Cc: valgrind-users@lists.sourceforge.net Subject: Re: [Valgrind-users] Uniformity of inter-snapshot times in

Re: [Valgrind-users] Uniformity of inter-snapshot times in Massif

2009-09-01 Thread Nicholas Nethercote
On Wed, Sep 2, 2009 at 3:20 AM, Belmon, Stephane wrote: > When Massif culls "old" snapshots, it drops about half of them, and then > resets its period to the MIN() of the inter-snapshots times. > > For long runs, that has somewhat unexpected consequences. You can easily end > up with 75 snapshots

[Valgrind-users] Uniformity of inter-snapshot times in Massif

2009-09-01 Thread Belmon, Stephane
When Massif culls "old" snapshots, it drops about half of them, and then resets its period to the MIN() of the inter-snapshots times. For long runs, that has somewhat unexpected consequences. You can easily end up with 75 snapshots that are about a minute apart, and then have the last 25 about