Bug 260155
| Summary: | inadequate LAYOUTTYPE sanity check in nfsv4_loadattr() | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | Base System | Reporter: | Robert Morris <rtm> | ||||||
| Component: | kern | Assignee: | Rick Macklem <rmacklem> | ||||||
| Status: | Closed FIXED | ||||||||
| Severity: | Affects Some People | CC: | emaste, rmacklem | ||||||
| Priority: | --- | Flags: | rmacklem:
mfc-stable13+
rmacklem: mfc-stable12+ |
||||||
| Version: | CURRENT | ||||||||
| Hardware: | Any | ||||||||
| OS: | Any | ||||||||
| Attachments: |
|
||||||||
Created attachment 229868 [details]
sanity check the count of layouttypes in attributes
This patch should fix the crash.
Maybe the reporter can confirm it fixes the problem for him?
(In reply to Rick Macklem from comment #1) The patch does fix the problem for me. A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=480be96e1e24617f0f152256d7453f60774fd1f3 commit 480be96e1e24617f0f152256d7453f60774fd1f3 Author: Rick Macklem <rmacklem@FreeBSD.org> AuthorDate: 2021-12-04 22:18:48 +0000 Commit: Rick Macklem <rmacklem@FreeBSD.org> CommitDate: 2021-12-04 22:22:19 +0000 nfsd: Sanity check the Layouttype count Reported by: rtm@lcs.mit.edu Tested by: rtm@lcs.mit.edu PR: 260155 MFC after: 2 weeks sys/fs/nfs/nfs_commonsubs.c | 9 +++++++++ 1 file changed, 9 insertions(+) The patch has been committed and will be MFC'd. A commit in branch stable/13 references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=fac8422a41e61847acd06f7dab9fab9cfc2f4a74 commit fac8422a41e61847acd06f7dab9fab9cfc2f4a74 Author: Rick Macklem <rmacklem@FreeBSD.org> AuthorDate: 2021-12-04 22:18:48 +0000 Commit: Rick Macklem <rmacklem@FreeBSD.org> CommitDate: 2021-12-18 01:32:47 +0000 nfsd: Sanity check the Layouttype count PR: 260155 (cherry picked from commit 480be96e1e24617f0f152256d7453f60774fd1f3) sys/fs/nfs/nfs_commonsubs.c | 9 +++++++++ 1 file changed, 9 insertions(+) A commit in branch stable/12 references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=5ad7804beb38fdc23f0f161484f5469e38c8fade commit 5ad7804beb38fdc23f0f161484f5469e38c8fade Author: Rick Macklem <rmacklem@FreeBSD.org> AuthorDate: 2021-12-04 22:18:48 +0000 Commit: Rick Macklem <rmacklem@FreeBSD.org> CommitDate: 2021-12-18 01:41:43 +0000 nfsd: Sanity check the Layouttype count PR: 260155 (cherry picked from commit 480be96e1e24617f0f152256d7453f60774fd1f3) sys/fs/nfs/nfs_commonsubs.c | 9 +++++++++ 1 file changed, 9 insertions(+) Patch has been committed and MFC'd. |
Created attachment 229829 [details] Crash an NFS server with a broken LAYOUTTYPE attribute. In this nfsv4_loadattr() code: case NFSATTRBIT_FSLAYOUTTYPE: case NFSATTRBIT_LAYOUTTYPE: NFSM_DISSECT(tl, u_int32_t *, NFSX_UNSIGNED); attrsum += NFSX_UNSIGNED; i = fxdr_unsigned(int, *tl); if (i > 0) { NFSM_DISSECT(tl, u_int32_t *, i * NFSX_UNSIGNED); If the RPC contains an i that is positive and big enough that i*NFSX_UNSIGNED is negative, the next attribute may experience a crash due to nd_dpos being wild. nfsm_dissect() and nfsm_dissct() are only able to prevent this if the siz is positive, not negative. I've attached a demo: # uname -a FreeBSD 14.0-CURRENT FreeBSD 14.0-CURRENT #120 main-n250906-d95bc6b0bf4c-dirty: Wed Dec 1 06:52:50 EST 2021 rtm@xxx:/usr/obj/usr/rtm/symbsd/src/riscv.riscv64/sys/RTM riscv # cc fnfsd_8.c # ./a.out ... panic: Fatal page fault at 0xffffffc000209adc: 0xffffffcf818ecbc0 --- exception 13, tval = 0xffffffcf818ecbc0 nfsv4_loadattr() at nfsv4_loadattr+0xef8 nfsrvd_verify() at nfsrvd_verify+0xb6 nfsrvd_dorpc() at nfsrvd_dorpc+0x147a nfssvc_program() at nfssvc_program+0x5a8 svc_run_internal() at svc_run_internal+0x810 svc_run() at svc_run+0x1a2 nfsrvd_nfsd() at nfsrvd_nfsd+0x30c nfssvc_nfsd() at nfssvc_nfsd+0x3ac sys_nfssvc() at sys_nfssvc+0xd0 do_trap_user() at do_trap_user+0x220 cpu_exception_handler_user() at cpu_exception_handler_user+0x72 --- exception 8, tval = 0x1c5816ef20