I solved my problem with xorg: it was indeed a bug in radeon driver, and I've verified that it's present both in unstable and experimental sources, but it should affect only people using radeon driver *with disabled acceleration*, which seems to be unrelated to the original notification.
If anyone has the possibility to debug its xserver and reproduce the segfault, the bug I found is in the function RADEONSetPortAttribute (file xc/programs/Xserver/hw/xfree86/driver/ati/radeon_video.c). The first statement reads: info->accel->Sync(pScrn); and should be modified into: if (info->accelOn) info->accel->Sync(pScrn); as info->accel, when info->accelOn is false, can be NULL or unmeaningful. -- Riccardo Brigo Scuola Superiore Sant'Anna - Pisa PGP Key available at http://pgp.mit.edu KEY FINGERPRINT: EB67 18FE 8891 5A19 B886 E6E0 6DDC 47EF E0FC 2681
signature.asc
Description: OpenPGP digital signature