[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: how to deal with 'trap invalid opcode'?



On Thu, Feb 13, 2020 at 08:03:03AM -0500, Henning Follmann wrote:
> On Thu, Feb 13, 2020 at 12:24:34PM +0100, Patrice Duroux wrote:
> > Hi,
> > 
> > The system on which it i occurring is a Debian Buster 10.3
> > The binaries that are involved are provided by the following packages:
> > ii  ncbi-blast+                               2.8.1-1
> >       amd64        next generation suite of BLAST sequence search tools
> > ii  ncbi-blast+-legacy                        2.8.1-1
> >       all          NCBI Blast legacy call script
> > 
> > and here is what dmesg says:
> > 
> > [16950090.751977] traps: blastn[8041] trap invalid opcode ip:7f1a5e4ce98c
> > sp:7ffe69e29be0 error:0 in libblast.so[7f1a5e48d000+66000]
> > [16951820.852031] traps: blastn[8512] trap invalid opcode ip:7fd312fed98c
> > sp:7ffe840c5780 error:0 in libblast.so[7fd312fac000+66000]
> > [16952207.586375] traps: makeblastdb[8966] trap invalid opcode
> > ip:7faa9aaa6b02 sp:7ffe4ff58040 error:0 in libxutil.so[7faa9aa73000+83000]
> > [16952218.407194] traps: makeblastdb[8975] trap invalid opcode
> > ip:7f8a2ac08b02 sp:7ffc88589580 error:0 in libxutil.so[7f8a2abd5000+83000]
> > 
> > May some amd64 packages supposed to be executable only for specific
> > CPU/arch?
> > 
> > Here is the hardware information:
> > 
> > LANG=C lshw -short
> > H/W path                Device      Class          Description
> > ==============================================================
> >                                     system         PowerEdge 6850
> > /0                                  bus            0RD318
> > /0/0                                memory         64KiB BIOS
> > /0/400                              processor      Intel(R) Xeon(TM) CPU
> > 3.40GHz
> > /0/400/700                          memory         32KiB L1 cache
> > /0/400/701                          memory         2MiB L2 cache
> > /0/400/702                          memory         16MiB L3 cache
> > /0/401                              processor      Intel(R) Xeon(TM) CPU
> > 3.40GHz
> > /0/401/703                          memory         32KiB L1 cache
> > /0/401/704                          memory         2MiB L2 cache
> > /0/401/705                          memory         16MiB L3 cache
> > /0/402                              processor      Intel(R) Xeon(TM) CPU
> > 3.40GHz
> > /0/402/706                          memory         32KiB L1 cache
> > /0/402/707                          memory         2MiB L2 cache
> > /0/402/708                          memory         16MiB L3 cache
> > /0/403                              processor      Intel(R) Xeon(TM) CPU
> > 3.40GHz
> > /0/403/709                          memory         32KiB L1 cache
> > /0/403/70a                          memory         2MiB L2 cache
> > /0/403/70b                          memory         16MiB L3 cache
> > /0/1000                             memory         64GiB System Memory
> > ...
> > 
> > Thanks,
> > Patrice
> 
> Your information is not helpful AT ALL.
> What are you are actually executing? What led to this error?
> Nobody has a chrystal bowl, being able to look  over your
> shoulder in the past (they are only for the future anyway).

It's not *that* bad.  We got the exact CPU model, the exact package
version, and the exact error message.

I'd say it's enough information to file a bug report with.  Adding
a recipe for how to produce the error messages would be good, too.
That's really the only missing piece.


Reply to: