| To: | Martin Michlmayr <tbm@cyrius.com>, linux-mips@linux-mips.org |
|---|---|
| Subject: | Re: [PATCH] gettimeofday jumps backwards then forwards |
| From: | Dave Johnson <djohnson+linuxmips@sw.starentnetworks.com> |
| Date: | Mon, 16 Jan 2006 11:35:26 -0500 |
| In-reply-to: | <20060116160031.GA28383@deprecation.cyrius.com> |
| Original-recipient: | rfc822;linux-mips@linux-mips.org |
| References: | <17118.25343.948383.547250@cortez.sw.starentnetworks.com> <20060116160031.GA28383@deprecation.cyrius.com> |
| Sender: | linux-mips-bounce@linux-mips.org |
Martin Michlmayr writes: > * Dave Johnson <djohnson+linuxmips@sw.starentnetworks.com> [2005-07-20 10:43]: > > Below are 2 fixes I made to 2.6.12 to do with time jumping around > > as reported by gettimeofday(). One is SB1250 specific and one appears > > generic. > > > > The symptom is revealed by running multile copies (1 per cpu) of a > > simple test program that calls gettimeofday() as fast as possible > > looking for time to go backwards. > > > > When a jump is detected the program outputs a few samples before and > > after each jump: > > Does anyone have comments regarding this patch? In addition to this problem I found significant lost ticks under load as there is no checking for lost ticks. I'll create a clean patch with just that fix in a bit. -- Dave Johnson Starent Networks |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH Cobalt 1/1] 64-bit fix, Jim Gifford |
|---|---|
| Next by Date: | Re: [PATCH Cobalt 1/1] 64-bit fix, Martin Michlmayr |
| Previous by Thread: | Re: [PATCH] gettimeofday jumps backwards then forwards, Martin Michlmayr |
| Next by Thread: | [PATCH] fix lost ticks on SB1250 [WAS: Re: [PATCH] gettimeofday jumps backwards then forwards], Dave Johnson |
| Indexes: | [Date] [Thread] [Top] [All Lists] |