Date: Mon, 10 May 2010 12:46:21 +0300 From: Achilleas Mantzios <achill@matrix.gatewaynet.com> To: freebsd-java@freebsd.org Subject: Re: openjdk6 crashes Message-ID: <201005101246.22215.achill@matrix.gatewaynet.com> In-Reply-To: <4BE4843E.6030901@gmail.com> References: <4BE11114.5030306@gmail.com> <201005070902.22700.achill@matrix.gatewaynet.com> <4BE4843E.6030901@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
=D3=F4=E9=F2 Saturday 08 May 2010 00:21:02 =EF/=E7 Peter Palmreuther =DD=E3= =F1=E1=F8=E5: >=20 > *erm* Maybe. For fixed conditions (which are seldom in my experience). Mo= st > times I've seen systems crashing here and there with no fixed script or > program if hardware dies. Simply because it's not deterministic what > temperature the system will have and what program will be at which memory > address. The only fixed crashing point can be the hard disk, which I thin= k I > can exclude due to the recompilation and therefore "relocation" on disk (= i.e. > different inodes being used). it has happened to me with my home FreeBSD box, crashing every morning at 0= 3:00, after a system upgrade (i think it was from 6.2->7.1). The easy answer was to accuse the software (upgrade), however what was most= probably happening was that the system upgrade made a previously hidden hardware problem to ap= pear. The system did not came back to stability, untill i replaced the motherboar= d. (after having tried changing the usual suspect RAM) The program causing the panic was periodic daily. =46or some reason, the same script on the same hardware with the previous 6= =2E2 kernel was working ok. >=20 > And to come to an end: if I see a program written in C++ (like OpenJDK/JV= M) > constantly fails on an assertion (like in my case: 'Error: assert(sp() >= =3D > inputs,"must have enough JVMS stack to execute")') I expect the assertion= to > not match first and think about corrupt memory a few hundred positions la= ter > on the "check list". That's at least what assertions are for: make the pr= ogram > fail if something is recognizable wrong. If every "assert fail" would be a > memory defect unless 100% proven not to be that would render assertions > useless, wouldn't it? >=20 > But thanks for your help and input; Maybe somebody else has any idea abou= t why > "sp()" is less than "input" when compiling (hotspot compiling?) this spec= ial > class in OpenJDK?!? Anyway Have your tried=20 =2D native jdk=20 =2D the binary diablo=20 =2D the linux jdk with linuxolator ?? =2D-=20 Achilleas Mantzios
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201005101246.22215.achill>