[QE-users] SCF calculation fails with SIGSEGV in Quantum ESPRESSO 7.5

[Paolo Giannozzi]

The error you get means one of the following things happened:

1. out of memory
I am not aware of any recent changes that may have increased the memory 
requirement. An estimate of the needed RAM is printed at the beginning 
of the calculation. On 8 processors, 1.55 Gb / processor, 12.4 Gb total.

2. something seriously wrong in the code
Very unlikely: your calculation is a standard one, I've tried it and it 
works (see below)

3. bad compilation, or bad libraries linked
Hard to say more without more info (and even with more info) on how the 
code was compiled,  how it was run, on which machine ...

By the way: "diagonalization='cg'" is horribly slow and should not be 
used any longer.

Paolo

On 3/30/26 07:20, Ishiyama,Takahisa_石山 貴久 wrote:
> Hi,
> 
> I am facing a problem regarding the scf calculation.
> 
> Scf calculation run successfully with version 7.4, but it fails with 
> version 7.5, even though I use essentially the same input.
> 
> To resolve this issue, I have tried changing some parameters , such as 
> smearing, degauss, cutoff, mixing_beta and so on.
> 
> However, the problem persists in version 7.5.
> 
> Coud this issue be related to changes introduced in version 7.5?
> 
> I am attaching the input file and the part of log below.
> 
> negative rho (up, down):  2.798E-05 0.000E+00
> 
> forrtl: severe (174): SIGSEGV, segmentation fault occurred
> 
> ===================================================================================
> 
> =   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
> 
> =   RANK 119 PID 19859 RUNNING AT 172.16.1.207
> 
> =   KILLED BY SIGNAL: 9 (Killed)
> 
> ===================================================================================
> 
> Exited with code 0
> 
> 
> _______________________________________________________________________________
> The Quantum ESPRESSO Foundation stands in solidarity with all civilians worldwide who are victims of terrorism, military aggression, and indiscriminate warfare.
> --------------------------------------------------------------------------------
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users

-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine Italy, +39-0432-558216