Re: [go-nuts] corrupt stack?

after upgrading to 1.9 (50% reduction) and finding a data race we didn't see in testing, we're still hunting down about 1 crash per 67 million hours of runtime. needless to say, the economical thing to do is to ignore the problem, but it sure bugs (!) me. the correct number of crashes is 0. -

Re: [go-nuts] corrupt stack?

i should have mentioned the race detector found nothing. jan, can you give an example of a go program setting a pointer to -1 without using unsafe? this requires the gc to have free'd something that is still live, doesn't it? - erik On Monday, December 4, 2017 at 2:58:22 AM UTC-8, Peter

Re: [go-nuts] corrupt stack?

a program not using unsafe cannot set a pointer to -1, even if there is a race, right? - erik On Sunday, December 3, 2017 at 10:20:44 PM UTC-8, Jan Mercl wrote: > > On Mon, Dec 4, 2017 at 7:03 AM wrote: > > > does anyone have any idea what's going on here, or some hints on

[go-nuts] corrupt stack?

i'm running go 1.8.3 on linux. about 1 in ~10 billion calls (spread across many machines), i get a backtrace that looks like the following: panic: runtime error: invalid memory address or nil pointer dereference [signal SIGSEGV: segmentation violation code=0x1 addr=0x28 pc=0x4c3406] goroutine