| To: | "'Frank Ch. Eigler'" <fche@xxxxxxxxxx> |
|---|---|
| Subject: | RE: pcp+graphite, take 2 |
| From: | "Ken McDonell" <kenj@xxxxxxxxxxxxxxxx> |
| Date: | Fri, 20 Jun 2014 15:08:49 +1000 |
| Cc: | <pcp@xxxxxxxxxxx> |
| Delivered-to: | pcp@xxxxxxxxxxx |
| In-reply-to: | <y0megylufs6.fsf@xxxxxxxx> |
| References: | <20140616214301.GB6693@xxxxxxxxxx> <539F7C0B.9050507@xxxxxxxxxxxxxxxx> <53A2297A.6000202@xxxxxxxxxxxxxxxx> <y0megylufs6.fsf@xxxxxxxx> |
| Thread-index: | AQMFYlM3q7i8zsnFcP5mMmn8PppB7AD8Ejm3AxKfV0UBsloF25jfvRow |
Frank, Just a quick update ... I have pmwebd working with -M 4 now ... but it required a lobotomy for the read_cache used in interp.c ... given the risks here and the complexity of the code, don't expect a commit "any time real soon", but don't waste time trying to debug in parallel either. Initial speedup is only 10% on a 2 CPU VM ... so that's not too encouraging ... but let's wait until I have the final fix done before we get too excited or too depressed. > -----Original Message----- > From: Frank Ch. Eigler [mailto:fche@xxxxxxxxxx] > Sent: Friday, 20 June 2014 12:37 AM > To: Ken McDonell > Cc: pcp@xxxxxxxxxxx > Subject: Re: pcp+graphite, take 2 > ... > (It turns out that -M1 etc. still crashes quickly, so we have some way to go > before multithreaded speedorama.) |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: PCP Network Latency PMDA, Frank Ch. Eigler |
|---|---|
| Next by Date: | Re: [pcp] pmServiceDiscoveryInterrupt() commit a8b87e2 et al., Dave Brolley |
| Previous by Thread: | Re: pcp+graphite, take 2, Frank Ch. Eigler |
| Next by Thread: | pcp updates: gnu/hurd port, release prep, Nathan Scott |
| Indexes: | [Date] [Thread] [Top] [All Lists] |