possible bug in ipc/msg.c|
Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.arcetri.astro.it/irlab/doc/library/linux/khg/HyperNews/get/khg/230.html
Дата изменения: Thu Mar 23 12:37:54 2000 Дата индексирования: Wed Sep 15 00:51:51 2010 Кодировка: Поисковые слова: п п п п п п п п п п п п п п р п р п |
possible bug in ipc/msg.c
|
hi first, i hope that this is the right place, :-> , since i an not sure about the 'finding' ... i need an advice. i am currently reading the kernel's code systematiclay, and i believe i stumbled into a bug in ipc/msg.c lines 326,329. i am talking about kernel 2.0.30-2.0.33 ( which i am working with ) and not about the development kernels... please read the relevent code ... since we are no longer ( between this lines ) in atomic operations, someone can suspend are in say line 326, recieve the current message ( the one we have nmsg as pointer to ) and leave us with pointer to garbage... i belive that we should put lines 326-329 in cli/restoreflags() pair after checking that the message is valid via the pointer flag ( not IPC_UNUSED/IPC_NOOID ). i hope that i am not bothering you for nothing... i have a possible patch. thank you for your time Michael ( m_photon@usa.net ) |