Skip to content

[ROOT-2927] Particular instance of data causes segfault in TDecompSparse #14554

Open
@vepadulano

Description

@vepadulano

Check duplicate issues.

  • Checked for duplicates

Description

From https://its.cern.ch/jira/browse/ROOT-2927

Dear ROOT,

I am attaching to this message a ROOT file that contains a sparse matrix, and a vector.

Trying to solve the equation formed by the two causes a segfault. The steps to reproduce the problem are the following:

$ root \-l matrix_fail.root
root [0]
Attaching file matrix_fail.root as _file0...
root [1] M
(class TMatrixTSparse<double>*)0xbed850
root [2] x
(class TVectorT<double>*)0xbedcb0
root [3] TDecompSparse solver(*M,0)
root [4] Bool_t ok
root [5] TVectorD xx = solver.Solve(*x, ok)

*** Break *** segmentation violation

The problem is due to this particular instance of the data: if you change one of the values of the x vector for example, you won't observe the segfault.

I have tried this in two different linux distributions (SL5 and ubuntu karmic) and two different ROOT versions (5.24 and 5.18), both 64 bits.

Reproducer

As per the description

ROOT version

Any

Installation method

Any

Operating system

Any

Additional context

Find input file in the Jira issue

Metadata

Metadata

Assignees

Type

No type

Projects

Status

Stuck - need help

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions