Re: Physical page zeroing on allocation

2013-04-03 Thread Chow Loong Jin
On 04/04/2013 06:39, Viacheslav Fedorov wrote: > Samuel Thibault mailto:sthiba...@debian.org>> writes: >>>when allocating an anonymous physical page on a Page Fault, >>> why does kernel have to fill it with zeroes? >>> I understand it has something to do with security. >>See man mmap. Anonymous mem

Re: Physical page zeroing on allocation

2013-04-03 Thread Viacheslav Fedorov
Samuel Thibault writes: >>when allocating an anonymous physical page on a Page Fault, >> why does kernel have to fill it with zeroes? >> I understand it has something to do with security. >See man mmap. Anonymous memory is explicitly documented as being set to >0. >Samuel Ok,thanks. So libraries

Re: Physical page zeroing on allocation

2013-04-01 Thread Timo Juhani Lindfors
Viacheslav Fedorov writes: > But let's say I trust my environment. Then why when I disable zeroing of > the allocated pages, the applications crash in libc? How did you disable it? -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Conta

Re: Physical page zeroing on allocation

2013-04-01 Thread Samuel Thibault
Viacheslav Fedorov, le Mon 01 Apr 2013 11:11:44 -0500, a écrit : > I was wondering, when allocating an anonymous physical page on a Page Fault, > why does kernel have to fill it with zeroes? > I understand it has something to do with security. > But let's say I trust my environment. Then why when I