From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org,
Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
Borislav Petkov <bp@suse.de>, Rik van Riel <riel@surriel.com>,
Aubrey Li <aubrey.li@intel.com>,
Austin Clements <austin@google.com>,
Barret Rhoden <brho@google.com>,
Dave Hansen <dave.hansen@intel.com>,
David Chase <drchase@golang.org>,
"H. Peter Anvin" <hpa@zytor.com>,
ian@airs.com, Ingo Molnar <mingo@redhat.com>,
Josh Bleecher Snyder <josharian@gmail.com>,
Thomas Gleixner <tglx@linutronix.de>, x86-ml <x86@kernel.org>
Subject: [PATCH 5.4 15/46] x86/fpu: Dont cache access to fpu_fpregs_owner_ctx
Date: Tue, 3 Dec 2019 23:35:35 +0100 [thread overview]
Message-ID: <20191203212732.978208719@linuxfoundation.org> (raw)
In-Reply-To: <20191203212705.175425505@linuxfoundation.org>
From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
commit 59c4bd853abcea95eccc167a7d7fd5f1a5f47b98 upstream.
The state/owner of the FPU is saved to fpu_fpregs_owner_ctx by pointing
to the context that is currently loaded. It never changed during the
lifetime of a task - it remained stable/constant.
After deferred FPU registers loading until return to userland was
implemented, the content of fpu_fpregs_owner_ctx may change during
preemption and must not be cached.
This went unnoticed for some time and was now noticed, in particular
since gcc 9 is caching that load in copy_fpstate_to_sigframe() and
reusing it in the retry loop:
copy_fpstate_to_sigframe()
load fpu_fpregs_owner_ctx and save on stack
fpregs_lock()
copy_fpregs_to_sigframe() /* failed */
fpregs_unlock()
*** PREEMPTION, another uses FPU, changes fpu_fpregs_owner_ctx ***
fault_in_pages_writeable() /* succeed, retry */
fpregs_lock()
__fpregs_load_activate()
fpregs_state_valid() /* uses fpu_fpregs_owner_ctx from stack */
copy_fpregs_to_sigframe() /* succeeds, random FPU content */
This is a comparison of the assembly produced by gcc 9, without vs with this
patch:
| # arch/x86/kernel/fpu/signal.c:173: if (!access_ok(buf, size))
| cmpq %rdx, %rax # tmp183, _4
| jb .L190 #,
|-# arch/x86/include/asm/fpu/internal.h:512: return fpu == this_cpu_read_stable(fpu_fpregs_owner_ctx) && cpu == fpu->last_cpu;
|-#APP
|-# 512 "arch/x86/include/asm/fpu/internal.h" 1
|- movq %gs:fpu_fpregs_owner_ctx,%rax #, pfo_ret__
|-# 0 "" 2
|-#NO_APP
|- movq %rax, -88(%rbp) # pfo_ret__, %sfp
…
|-# arch/x86/include/asm/fpu/internal.h:512: return fpu == this_cpu_read_stable(fpu_fpregs_owner_ctx) && cpu == fpu->last_cpu;
|- movq -88(%rbp), %rcx # %sfp, pfo_ret__
|- cmpq %rcx, -64(%rbp) # pfo_ret__, %sfp
|+# arch/x86/include/asm/fpu/internal.h:512: return fpu == this_cpu_read(fpu_fpregs_owner_ctx) && cpu == fpu->last_cpu;
|+#APP
|+# 512 "arch/x86/include/asm/fpu/internal.h" 1
|+ movq %gs:fpu_fpregs_owner_ctx(%rip),%rax # fpu_fpregs_owner_ctx, pfo_ret__
|+# 0 "" 2
|+# arch/x86/include/asm/fpu/internal.h:512: return fpu == this_cpu_read(fpu_fpregs_owner_ctx) && cpu == fpu->last_cpu;
|+#NO_APP
|+ cmpq %rax, -64(%rbp) # pfo_ret__, %sfp
Use this_cpu_read() instead this_cpu_read_stable() to avoid caching of
fpu_fpregs_owner_ctx during preemption points.
The Fixes: tag points to the commit where deferred FPU loading was
added. Since this commit, the compiler is no longer allowed to move the
load of fpu_fpregs_owner_ctx somewhere else / outside of the locked
section. A task preemption will change its value and stale content will
be observed.
[ bp: Massage. ]
Debugged-by: Austin Clements <austin@google.com>
Debugged-by: David Chase <drchase@golang.org>
Debugged-by: Ian Lance Taylor <ian@airs.com>
Fixes: 5f409e20b7945 ("x86/fpu: Defer FPU state load until return to userspace")
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Rik van Riel <riel@surriel.com>
Tested-by: Borislav Petkov <bp@suse.de>
Cc: Aubrey Li <aubrey.li@intel.com>
Cc: Austin Clements <austin@google.com>
Cc: Barret Rhoden <brho@google.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Chase <drchase@golang.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: ian@airs.com
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Josh Bleecher Snyder <josharian@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: x86-ml <x86@kernel.org>
Link: https://lkml.kernel.org/r/20191128085306.hxfa2o3knqtu4wfn@linutronix.de
Link: https://bugzilla.kernel.org/show_bug.cgi?id=205663
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
arch/x86/include/asm/fpu/internal.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/arch/x86/include/asm/fpu/internal.h
+++ b/arch/x86/include/asm/fpu/internal.h
@@ -509,7 +509,7 @@ static inline void __fpu_invalidate_fpre
static inline int fpregs_state_valid(struct fpu *fpu, unsigned int cpu)
{
- return fpu == this_cpu_read_stable(fpu_fpregs_owner_ctx) && cpu == fpu->last_cpu;
+ return fpu == this_cpu_read(fpu_fpregs_owner_ctx) && cpu == fpu->last_cpu;
}
/*
next prev parent reply other threads:[~2019-12-03 22:39 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-03 22:35 [PATCH 5.4 00/46] 5.4.2-stable review Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 01/46] io_uring: async workers should inherit the user creds Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 02/46] net: separate out the msghdr copy from ___sys_{send,recv}msg() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 03/46] net: disallow ancillary data for __sys_{send,recv}msg_file() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 04/46] crypto: inside-secure - Fix stability issue with Macchiatobin Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 05/46] driver core: platform: use the correct callback type for bus_find_device Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 06/46] usb: dwc2: use a longer core rest timeout in dwc2_core_reset() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 07/46] staging: wilc1000: fix illegal memory access in wilc_parse_join_bss_param() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 08/46] staging: rtl8192e: fix potential use after free Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 09/46] staging: rtl8723bs: Drop ACPI device ids Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 10/46] staging: rtl8723bs: Add 024c:0525 to the list of SDIO device-ids Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 11/46] USB: serial: ftdi_sio: add device IDs for U-Blox C099-F9P Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 12/46] mei: bus: prefix device names on bus with the bus name Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 13/46] mei: me: add comet point V device id Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 14/46] thunderbolt: Power cycle the router if NVM authentication fails Greg Kroah-Hartman
2019-12-03 22:35 ` Greg Kroah-Hartman [this message]
2019-12-03 22:35 ` [PATCH 5.4 16/46] gve: Fix the queue page list allocated pages count Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 17/46] macvlan: schedule bc_work even if error Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 18/46] mdio_bus: dont use managed reset-controller Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 19/46] net: dsa: sja1105: fix sja1105_parse_rgmii_delays() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 20/46] net: macb: add missed tasklet_kill Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 21/46] net: psample: fix skb_over_panic Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 22/46] net: sched: fix `tc -s class show` no bstats on class with nolock subqueues Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 23/46] openvswitch: fix flow command message size Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 24/46] sctp: Fix memory leak in sctp_sf_do_5_2_4_dupcook Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 25/46] slip: Fix use-after-free Read in slip_open Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 26/46] sctp: cache netns in sctp_ep_common Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 27/46] openvswitch: drop unneeded BUG_ON() in ovs_flow_cmd_build_info() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 28/46] openvswitch: remove another BUG_ON() Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 29/46] net/tls: take into account that bpf_exec_tx_verdict() may free the record Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 30/46] net/tls: free the record on encryption error Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 31/46] net: skmsg: fix TLS 1.3 crash with full sk_msg Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 32/46] selftests/tls: add a test for fragmented messages Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 33/46] net/tls: remove the dead inplace_crypto code Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 34/46] net/tls: use sg_next() to walk sg entries Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 35/46] selftests: bpf: test_sockmap: handle file creation failures gracefully Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 36/46] selftests: bpf: correct perror strings Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 37/46] tipc: fix link name length check Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 38/46] selftests: pmtu: use -oneline for ip route list cache Greg Kroah-Hartman
2019-12-03 22:35 ` [PATCH 5.4 39/46] r8169: fix jumbo configuration for RTL8168evl Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 40/46] r8169: fix resume on cable plug-in Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 41/46] ext4: add more paranoia checking in ext4_expand_extra_isize handling Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 42/46] Revert "jffs2: Fix possible null-pointer dereferences in jffs2_add_frag_to_fragtree()" Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 43/46] crypto: talitos - Fix build error by selecting LIB_DES Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 44/46] HID: core: check whether Usage Page item is after Usage ID items Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 45/46] platform/x86: hp-wmi: Fix ACPI errors caused by too small buffer Greg Kroah-Hartman
2019-12-03 22:36 ` [PATCH 5.4 46/46] platform/x86: hp-wmi: Fix ACPI errors caused by passing 0 as input size Greg Kroah-Hartman
2019-12-04 10:26 ` [PATCH 5.4 00/46] 5.4.2-stable review Jon Hunter
2019-12-04 13:23 ` Amol Grover
2019-12-04 17:13 ` Greg Kroah-Hartman
2019-12-05 16:43 ` Amol Grover
2019-12-06 13:05 ` Greg Kroah-Hartman
2019-12-04 13:56 ` Naresh Kamboju
2019-12-04 20:38 ` Greg Kroah-Hartman
2019-12-04 17:50 ` shuah
2019-12-04 20:37 ` Greg Kroah-Hartman
2019-12-04 19:05 ` Guenter Roeck
2019-12-04 20:37 ` Greg Kroah-Hartman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191203212732.978208719@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=aubrey.li@intel.com \
--cc=austin@google.com \
--cc=bigeasy@linutronix.de \
--cc=bp@suse.de \
--cc=brho@google.com \
--cc=dave.hansen@intel.com \
--cc=drchase@golang.org \
--cc=hpa@zytor.com \
--cc=ian@airs.com \
--cc=josharian@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=riel@surriel.com \
--cc=stable@vger.kernel.org \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).