From 52d0b8b18776f184c53632c5e0068201491cdb61 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 26 Nov 2021 13:47:46 +0100 Subject: x86/fpu/signal: Initialize sw_bytes in save_xstate_epilog() save_sw_bytes() did not fully initialize sw_bytes, which caused KMSAN to report an infoleak (see below). Initialize sw_bytes explicitly to avoid this. KMSAN report follows: ===================================================== BUG: KMSAN: kernel-infoleak in instrument_copy_to_user ./include/linux/instrumented.h:121 BUG: KMSAN: kernel-infoleak in __copy_to_user ./include/linux/uaccess.h:154 BUG: KMSAN: kernel-infoleak in save_xstate_epilog+0x2df/0x510 arch/x86/kernel/fpu/signal.c:127 instrument_copy_to_user ./include/linux/instrumented.h:121 __copy_to_user ./include/linux/uaccess.h:154 save_xstate_epilog+0x2df/0x510 arch/x86/kernel/fpu/signal.c:127 copy_fpstate_to_sigframe+0x861/0xb60 arch/x86/kernel/fpu/signal.c:245 get_sigframe+0x656/0x7e0 arch/x86/kernel/signal.c:296 __setup_rt_frame+0x14d/0x2a60 arch/x86/kernel/signal.c:471 setup_rt_frame arch/x86/kernel/signal.c:781 handle_signal arch/x86/kernel/signal.c:825 arch_do_signal_or_restart+0x417/0xdd0 arch/x86/kernel/signal.c:870 handle_signal_work kernel/entry/common.c:149 exit_to_user_mode_loop+0x1f6/0x490 kernel/entry/common.c:173 exit_to_user_mode_prepare kernel/entry/common.c:208 __syscall_exit_to_user_mode_work kernel/entry/common.c:290 syscall_exit_to_user_mode+0x7e/0xc0 kernel/entry/common.c:302 do_syscall_64+0x60/0xd0 arch/x86/entry/common.c:88 entry_SYSCALL_64_after_hwframe+0x44/0xae ??:? Local variable sw_bytes created at: save_xstate_epilog+0x80/0x510 arch/x86/kernel/fpu/signal.c:121 copy_fpstate_to_sigframe+0x861/0xb60 arch/x86/kernel/fpu/signal.c:245 Bytes 20-47 of 48 are uninitialized Memory access of size 48 starts at ffff8880801d3a18 Data copied to user address 00007ffd90e2ef50 ===================================================== Link: https://lore.kernel.org/all/CAG_fn=V9T6OKPonSjsi9PmWB0hMHFC=yawozdft8i1-MSxrv=w@mail.gmail.com/ Fixes: 53599b4d54b9b8dd ("x86/fpu/signal: Prepare for variable sigframe length") Reported-by: Alexander Potapenko Signed-off-by: Marco Elver Signed-off-by: Alexander Potapenko Signed-off-by: Dave Hansen Tested-by: Alexander Potapenko Link: https://lkml.kernel.org/r/20211126124746.761278-1-glider@google.com --- arch/x86/kernel/fpu/signal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c index d5958278eba6..91d4b6de58ab 100644 --- a/arch/x86/kernel/fpu/signal.c +++ b/arch/x86/kernel/fpu/signal.c @@ -118,7 +118,7 @@ static inline bool save_xstate_epilog(void __user *buf, int ia32_frame, struct fpstate *fpstate) { struct xregs_state __user *x = buf; - struct _fpx_sw_bytes sw_bytes; + struct _fpx_sw_bytes sw_bytes = {}; u32 xfeatures; int err; -- cgit v1.3.1 From c7719e79347803b8e3b6b50da8c6db410a3012b5 Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Wed, 17 Nov 2021 10:37:50 +0800 Subject: x86/tsc: Add a timer to make sure TSC_adjust is always checked The TSC_ADJUST register is checked every time a CPU enters idle state, but Thomas Gleixner mentioned there is still a caveat that a system won't enter idle [1], either because it's too busy or configured purposely to not enter idle. Setup a periodic timer (every 10 minutes) to make sure the check is happening on a regular base. [1] https://lore.kernel.org/lkml/875z286xtk.fsf@nanos.tec.linutronix.de/ Fixes: 6e3cd95234dc ("x86/hpet: Use another crystalball to evaluate HPET usability") Requested-by: Thomas Gleixner Signed-off-by: Feng Tang Signed-off-by: Thomas Gleixner Cc: "Paul E. McKenney" Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20211117023751.24190-1-feng.tang@intel.com --- arch/x86/kernel/tsc_sync.c | 41 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 41 insertions(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index 50a4515fe0ad..9452dc9664b5 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -30,6 +30,7 @@ struct tsc_adjust { }; static DEFINE_PER_CPU(struct tsc_adjust, tsc_adjust); +static struct timer_list tsc_sync_check_timer; /* * TSC's on different sockets may be reset asynchronously. @@ -77,6 +78,46 @@ void tsc_verify_tsc_adjust(bool resume) } } +/* + * Normally the tsc_sync will be checked every time system enters idle + * state, but there is still caveat that a system won't enter idle, + * either because it's too busy or configured purposely to not enter + * idle. + * + * So setup a periodic timer (every 10 minutes) to make sure the check + * is always on. + */ + +#define SYNC_CHECK_INTERVAL (HZ * 600) + +static void tsc_sync_check_timer_fn(struct timer_list *unused) +{ + int next_cpu; + + tsc_verify_tsc_adjust(false); + + /* Run the check for all onlined CPUs in turn */ + next_cpu = cpumask_next(raw_smp_processor_id(), cpu_online_mask); + if (next_cpu >= nr_cpu_ids) + next_cpu = cpumask_first(cpu_online_mask); + + tsc_sync_check_timer.expires += SYNC_CHECK_INTERVAL; + add_timer_on(&tsc_sync_check_timer, next_cpu); +} + +static int __init start_sync_check_timer(void) +{ + if (!cpu_feature_enabled(X86_FEATURE_TSC_ADJUST) || tsc_clocksource_reliable) + return 0; + + timer_setup(&tsc_sync_check_timer, tsc_sync_check_timer_fn, 0); + tsc_sync_check_timer.expires = jiffies + SYNC_CHECK_INTERVAL; + add_timer(&tsc_sync_check_timer); + + return 0; +} +late_initcall(start_sync_check_timer); + static void tsc_sanitize_first_cpu(struct tsc_adjust *cur, s64 bootval, unsigned int cpu, bool bootcpu) { -- cgit v1.3.1 From b50db7095fe002fa3e16605546cba66bf1b68a3e Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Wed, 17 Nov 2021 10:37:51 +0800 Subject: x86/tsc: Disable clocksource watchdog for TSC on qualified platorms There are cases that the TSC clocksource is wrongly judged as unstable by the clocksource watchdog mechanism which tries to validate the TSC against HPET, PM_TIMER or jiffies. While there is hardly a general reliable way to check the validity of a watchdog, Thomas Gleixner proposed [1]: "I'm inclined to lift that requirement when the CPU has: 1) X86_FEATURE_CONSTANT_TSC 2) X86_FEATURE_NONSTOP_TSC 3) X86_FEATURE_NONSTOP_TSC_S3 4) X86_FEATURE_TSC_ADJUST 5) At max. 4 sockets After two decades of horrors we're finally at a point where TSC seems to be halfway reliable and less abused by BIOS tinkerers. TSC_ADJUST was really key as we can now detect even small modifications reliably and the important point is that we can cure them as well (not pretty but better than all other options)." As feature #3 X86_FEATURE_NONSTOP_TSC_S3 only exists on several generations of Atom processorz, and is always coupled with X86_FEATURE_CONSTANT_TSC and X86_FEATURE_NONSTOP_TSC, skip checking it, and also be more defensive to use maximal 2 sockets. The check is done inside tsc_init() before registering 'tsc-early' and 'tsc' clocksources, as there were cases that both of them had been wrongly judged as unreliable. For more background of tsc/watchdog, there is a good summary in [2] [tglx} Update vs. jiffies: On systems where the only remaining clocksource aside of TSC is jiffies there is no way to make this work because that creates a circular dependency. Jiffies accuracy depends on not missing a periodic timer interrupt, which is not guaranteed. That could be detected by TSC, but as TSC is not trusted this cannot be compensated. The consequence is a circulus vitiosus which results in shutting down TSC and falling back to the jiffies clocksource which is even more unreliable. [1]. https://lore.kernel.org/lkml/87eekfk8bd.fsf@nanos.tec.linutronix.de/ [2]. https://lore.kernel.org/lkml/87a6pimt1f.ffs@nanos.tec.linutronix.de/ [ tglx: Refine comment and amend changelog ] Fixes: 6e3cd95234dc ("x86/hpet: Use another crystalball to evaluate HPET usability") Suggested-by: Thomas Gleixner Signed-off-by: Feng Tang Signed-off-by: Thomas Gleixner Cc: "Paul E. McKenney" Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20211117023751.24190-2-feng.tang@intel.com --- arch/x86/kernel/tsc.c | 28 ++++++++++++++++++++++++---- 1 file changed, 24 insertions(+), 4 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 2e076a459a0c..a698196377be 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -1180,6 +1180,12 @@ void mark_tsc_unstable(char *reason) EXPORT_SYMBOL_GPL(mark_tsc_unstable); +static void __init tsc_disable_clocksource_watchdog(void) +{ + clocksource_tsc_early.flags &= ~CLOCK_SOURCE_MUST_VERIFY; + clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY; +} + static void __init check_system_tsc_reliable(void) { #if defined(CONFIG_MGEODEGX1) || defined(CONFIG_MGEODE_LX) || defined(CONFIG_X86_GENERIC) @@ -1196,6 +1202,23 @@ static void __init check_system_tsc_reliable(void) #endif if (boot_cpu_has(X86_FEATURE_TSC_RELIABLE)) tsc_clocksource_reliable = 1; + + /* + * Disable the clocksource watchdog when the system has: + * - TSC running at constant frequency + * - TSC which does not stop in C-States + * - the TSC_ADJUST register which allows to detect even minimal + * modifications + * - not more than two sockets. As the number of sockets cannot be + * evaluated at the early boot stage where this has to be + * invoked, check the number of online memory nodes as a + * fallback solution which is an reasonable estimate. + */ + if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC) && + boot_cpu_has(X86_FEATURE_NONSTOP_TSC) && + boot_cpu_has(X86_FEATURE_TSC_ADJUST) && + nr_online_nodes <= 2) + tsc_disable_clocksource_watchdog(); } /* @@ -1387,9 +1410,6 @@ static int __init init_tsc_clocksource(void) if (tsc_unstable) goto unreg; - if (tsc_clocksource_reliable || no_tsc_watchdog) - clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY; - if (boot_cpu_has(X86_FEATURE_NONSTOP_TSC_S3)) clocksource_tsc.flags |= CLOCK_SOURCE_SUSPEND_NONSTOP; @@ -1527,7 +1547,7 @@ void __init tsc_init(void) } if (tsc_clocksource_reliable || no_tsc_watchdog) - clocksource_tsc_early.flags &= ~CLOCK_SOURCE_MUST_VERIFY; + tsc_disable_clocksource_watchdog(); clocksource_register_khz(&clocksource_tsc_early, tsc_khz); detect_art(); -- cgit v1.3.1 From 1d5379d0475419085d3575bd9155f2e558e96390 Mon Sep 17 00:00:00 2001 From: Michael Sterritt Date: Fri, 19 Nov 2021 15:27:57 -0800 Subject: x86/sev: Fix SEV-ES INS/OUTS instructions for word, dword, and qword Properly type the operands being passed to __put_user()/__get_user(). Otherwise, these routines truncate data for dependent instructions (e.g., INSW) and only read/write one byte. This has been tested by sending a string with REP OUTSW to a port and then reading it back in with REP INSW on the same port. Previous behavior was to only send and receive the first char of the size. For example, word operations for "abcd" would only read/write "ac". With change, the full string is now written and read back. Fixes: f980f9c31a923 (x86/sev-es: Compile early handler code into kernel image) Signed-off-by: Michael Sterritt Signed-off-by: Borislav Petkov Reviewed-by: Paolo Bonzini Reviewed-by: Marc Orr Reviewed-by: Peter Gonda Reviewed-by: Joerg Roedel Link: https://lkml.kernel.org/r/20211119232757.176201-1-sterritt@google.com --- arch/x86/kernel/sev.c | 57 +++++++++++++++++++++++++++++++++++---------------- 1 file changed, 39 insertions(+), 18 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/sev.c b/arch/x86/kernel/sev.c index 74f0ec955384..a9fc2ac7a8bd 100644 --- a/arch/x86/kernel/sev.c +++ b/arch/x86/kernel/sev.c @@ -294,11 +294,6 @@ static enum es_result vc_write_mem(struct es_em_ctxt *ctxt, char *dst, char *buf, size_t size) { unsigned long error_code = X86_PF_PROT | X86_PF_WRITE; - char __user *target = (char __user *)dst; - u64 d8; - u32 d4; - u16 d2; - u8 d1; /* * This function uses __put_user() independent of whether kernel or user @@ -320,26 +315,42 @@ static enum es_result vc_write_mem(struct es_em_ctxt *ctxt, * instructions here would cause infinite nesting. */ switch (size) { - case 1: + case 1: { + u8 d1; + u8 __user *target = (u8 __user *)dst; + memcpy(&d1, buf, 1); if (__put_user(d1, target)) goto fault; break; - case 2: + } + case 2: { + u16 d2; + u16 __user *target = (u16 __user *)dst; + memcpy(&d2, buf, 2); if (__put_user(d2, target)) goto fault; break; - case 4: + } + case 4: { + u32 d4; + u32 __user *target = (u32 __user *)dst; + memcpy(&d4, buf, 4); if (__put_user(d4, target)) goto fault; break; - case 8: + } + case 8: { + u64 d8; + u64 __user *target = (u64 __user *)dst; + memcpy(&d8, buf, 8); if (__put_user(d8, target)) goto fault; break; + } default: WARN_ONCE(1, "%s: Invalid size: %zu\n", __func__, size); return ES_UNSUPPORTED; @@ -362,11 +373,6 @@ static enum es_result vc_read_mem(struct es_em_ctxt *ctxt, char *src, char *buf, size_t size) { unsigned long error_code = X86_PF_PROT; - char __user *s = (char __user *)src; - u64 d8; - u32 d4; - u16 d2; - u8 d1; /* * This function uses __get_user() independent of whether kernel or user @@ -388,26 +394,41 @@ static enum es_result vc_read_mem(struct es_em_ctxt *ctxt, * instructions here would cause infinite nesting. */ switch (size) { - case 1: + case 1: { + u8 d1; + u8 __user *s = (u8 __user *)src; + if (__get_user(d1, s)) goto fault; memcpy(buf, &d1, 1); break; - case 2: + } + case 2: { + u16 d2; + u16 __user *s = (u16 __user *)src; + if (__get_user(d2, s)) goto fault; memcpy(buf, &d2, 2); break; - case 4: + } + case 4: { + u32 d4; + u32 __user *s = (u32 __user *)src; + if (__get_user(d4, s)) goto fault; memcpy(buf, &d4, 4); break; - case 8: + } + case 8: { + u64 d8; + u64 __user *s = (u64 __user *)src; if (__get_user(d8, s)) goto fault; memcpy(buf, &d8, 8); break; + } default: WARN_ONCE(1, "%s: Invalid size: %zu\n", __func__, size); return ES_UNSUPPORTED; -- cgit v1.3.1