s390/pci: handle insufficient resources during dma tlb flush

In a virtualized setup lazy flushing can lead to the hypervisor
running out of resources when lots of guest pages need to be
pinned. In this situation simply trigger a global flush to give
the hypervisor a chance to free some of these resources.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Reviewed-by: Pierre Morel <pmorel@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This commit is contained in:
Sebastian Ott
2017-12-01 18:47:32 +01:00
committed by Martin Schwidefsky
parent d39a01eff9
commit a5f1005517
2 changed files with 22 additions and 2 deletions

View File

@@ -89,6 +89,9 @@ int zpci_refresh_trans(u64 fn, u64 addr, u64 range)
if (cc)
zpci_err_insn(cc, status, addr, range);
if (cc == 1 && (status == 4 || status == 16))
return -ENOMEM;
return (cc) ? -EIO : 0;
}