smallsieve failure causes renumber to crash.
Imported issue: Initially reported by @thome in https://gforge.inria.fr/tracker/?group_id=2065&aid=21553
https://ci.inria.fr/cado/job/future-I18/job/compile-debian-testing-amd64-abi32-c
90/103/console
Error:Filtering - Duplicate Removal, removal pass: Command line was: /localdisk/ci/jenkins/workspace/future-I18/compile-debian-testing-amd64-abi32-c90/build/tartiflette/filter/dup2 -nrels 434535 -renumber /tmp/cado.arm5er5n/c90.renumber.gz /tmp/cado.arm5er5n/c90.dup1//1/dup1.0.0000.gz > /tmp/cado.arm5er5n/c90.dup2.slice1.stdout.1 2> /tmp/cado.arm5er5n/c90.dup2.slice1.stderr.1
Error:Filtering - Duplicate Removal, removal pass: Stderr output follows (stored in file /tmp/cado.arm5er5n/c90.dup2.slice1.stderr.1):
b'antebuffer set to /localdisk/ci/jenkins/workspace/future-I18/compile-debian-testing-amd64-abi32-c90/build/tartiflette/utils/antebuffer\n[checking true duplicates on sample of 5216 cells]\nAllocated hash table of 521551 entries (1Mb)\nConstructing the two filelists...\n1 files (1 new and 0 already renumbered)\nReading files already renumbered:\nReading new files (using 4 auxiliary threads for roots mod p):\nFatal error in renumber_get_first_index_from_p at /localdisk/ci/jenkins/workspace/future-I18/compile-debian-testing-amd64-abi32-c90/utils/renumber.c:899\nError with the cached part of the renumbering table\n p = 0x559\n vp = 0x55a\n i = cached[p] = 0x0\n tab[i] = 0x3\n'
Traceback (most recent call last):
File ""./cado-nfs.py"", line 122, in
factors = factorjob.run()
File ""./scripts/cadofactor/cadotask.py"", line 5436, in run
last_status, last_task = self.run_next_task()
File ""./scripts/cadofactor/cadotask.py"", line 5511, in run_next_task
return [task.run(), task.title]
File ""./scripts/cadofactor/cadotask.py"", line 3321, in run
raise Exception(""Program failed"")
Exception: Program failed
Two bugs here.
The root cause is that las seems to have printed a relation with a non-prime 1311 (=11^3). So this is a smallsieve failure that I intend to fix.
Now the gracefulness of this failure is a bit questionable.
I wish the tools were capable to provide more useful diagnostics in that case.
attachment: polynomial, roots, las_tracek command line (tx Pierrick)