@@ -1140,11 +1140,11 @@ SYSCALL_DEFINE0(rt_sigreturn)
{
struct rt_sigframe __user *rt_sf;
struct pt_regs *regs = current_pt_regs();
+ int tm_restore = 0;
#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
struct ucontext __user *uc_transact;
unsigned long msr_hi;
unsigned long tmp;
- int tm_restore = 0;
#endif
/* Always make any pending restarted system calls return -EINTR */
current->restart_block.fn = do_no_restart_syscall;
@@ -1192,11 +1192,19 @@ SYSCALL_DEFINE0(rt_sigreturn)
goto bad;
}
}
- if (!tm_restore)
- /* Fall through, for non-TM restore */
+ if (!tm_restore) {
+ /*
+ * Unset regs->msr because ucontext MSR TS is not
+ * set, and recheckpoint was not called. This avoid
+ * hitting a TM Bad thing at RFID
+ */
+ regs->msr &= ~MSR_TS_MASK;
+ }
+ /* Fall through, for non-TM restore */
#endif
- if (do_setcontext(&rt_sf->uc, regs, 1))
- goto bad;
+ if (!tm_restore)
+ if (do_setcontext(&rt_sf->uc, regs, 1))
+ goto bad;
/*
* It's not clear whether or why it is desirable to save the
@@ -740,11 +740,23 @@ SYSCALL_DEFINE0(rt_sigreturn)
&uc_transact->uc_mcontext))
goto badframe;
}
- else
- /* Fall through, for non-TM restore */
#endif
- if (restore_sigcontext(current, NULL, 1, &uc->uc_mcontext))
- goto badframe;
+ /* Fall through, for non-TM restore */
+ if (!MSR_TM_ACTIVE(msr)) {
+ /*
+ * Unset MSR[TS] on the thread regs since MSR from user
+ * context does not have MSR active, and recheckpoint was
+ * not called since restore_tm_sigcontexts() was not called
+ * also.
+ *
+ * If not unsetting it, the code can RFID to userspace with
+ * MSR[TS] set, but without CPU in the proper state,
+ * causing a TM bad thing.
+ */
+ current->thread.regs->msr &= ~MSR_TS_MASK;
+ if (restore_sigcontext(current, NULL, 1, &uc->uc_mcontext))
+ goto badframe;
+ }
if (restore_altstack(&uc->uc_stack))
goto badframe;
There is a TM Bad Thing bug that can be caused when you return from a signal context in a suspended transaction but with ucontext MSR[TS] unset. This forces regs->msr[TS] to be set at syscall entrance (since the CPU state is transactional). It also calls treclaim() to flush the transaction state, which is done based on the live (mfmsr) MSR state. Since user context MSR[TS] is not set, then restore_tm_sigcontexts() is not called, thus, not executing recheckpoint, keeping the CPU state as not transactional. When calling rfid, SRR1 will have MSR[TS] set, but the CPU state is non transactional, causing the TM Bad Thing with the following stack: [ 33.862316] Bad kernel stack pointer 3fffd9dce3e0 at c00000000000c47c cpu 0x8: Vector: 700 (Program Check) at [c00000003ff7fd40] pc: c00000000000c47c: fast_exception_return+0xac/0xb4 lr: 00003fff865f442c sp: 3fffd9dce3e0 msr: 8000000102a03031 current = 0xc00000041f68b700 paca = 0xc00000000fb84800 softe: 0 irq_happened: 0x01 pid = 1721, comm = tm-signal-sigre Linux version 4.9.0-3-powerpc64le (debian-kernel@lists.debian.org) (gcc version 6.3.0 20170516 (Debian 6.3.0-18) ) #1 SMP Debian 4.9.30-2+deb9u2 (2017-06-26) WARNING: exception is not recoverable, can't continue The same problem happens on 32-bits signal handler, and the fix is very similar, if tm_recheckpoint() is not executed, then regs->msr[TS] should be zeroed. This patch also fixes a sparse warning related to lack of indentation when CONFIG_PPC_TRANSACTIONAL_MEM is set. Fixes: 2b0a576d15e0e ("powerpc: Add new transactional memory state to the signal context") CC: Stable <stable@vger.kernel.org> # 3.10+ Signed-off-by: Breno Leitao <leitao@debian.org> --- arch/powerpc/kernel/signal_32.c | 18 +++++++++++++----- arch/powerpc/kernel/signal_64.c | 20 ++++++++++++++++---- 2 files changed, 29 insertions(+), 9 deletions(-)