| Summary: | pax problem in linux-headers-2.6 | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Daniel Black (RETIRED) <dragonheart> |
| Component: | [OLD] Core system | Assignee: | Tim Yamin (RETIRED) <plasmaroo> |
| Status: | RESOLVED NEEDINFO | ||
| Severity: | normal | CC: | hardened, netmon, pageexec |
| Priority: | High | ||
| Version: | 2005.0 | ||
| Hardware: | All | ||
| OS: | Other | ||
| URL: | http://forums.grsecurity.net./viewtopic.php?p=4741#4741 | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: | system.h.diff | ||
|
Description
Daniel Black (RETIRED)
2005-06-03 01:20:51 UTC
libpcap-ringbuffer fixed Created attachment 60518 [details, diff]
system.h.diff
Patch looks good, thanks, what version of headers you need this applying to? Will 2.6.11 be sufficient? actually, upon further thinking using 'awx' is not the best solution either, it will produce RWE PT_LOAD segments, which is not nice (albeit it happens to work even under PaX as no code is actually executed from there). the best solution would be to simply not expose the alternative() macro to userland at all, after all there's noone in userland who would actually make use of it (the kernel itself has special processing on boot that does the insn substitution when run on the proper CPU). short of that, use 'aw' everywhere. confession - I don't know much about the asm alternate stuff at all. I was just going of the refurl. There seems to be some confusion as to what would be the right way of doing things, so if somebody can attach an updated patch or comment if the current patch is correct and reopen I'll gladly add this in. Thanks! as i suggested, use "aw" section flags in all 4 places, i.e., the attached patch gets half of that right, the other half should be changed from "awx" into "aw". |