sparc64: Fix register corruption in top-most kernel stack frame during boot.

Meelis Roos reported that kernels built with gcc-4.9 do not boot, we
eventually narrowed this down to only impacting machines using
UltraSPARC-III and derivitive cpus.

The crash happens right when the first user process is spawned:

[   54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004
[   54.451346]
[   54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96
[   54.666431] Call Trace:
[   54.698453]  [0000000000762f8c] panic+0xb0/0x224
[   54.759071]  [000000000045cf68] do_exit+0x948/0x960
[   54.823123]  [000000000042cbc0] fault_in_user_windows+0xe0/0x100
[   54.902036]  [0000000000404ad0] __handle_user_windows+0x0/0x10
[   54.978662] Press Stop-A (L1-A) to return to the boot prom
[   55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004

Further investigation showed that compiling only per_cpu_patch() with
an older compiler fixes the boot.

Detailed analysis showed that the function is not being miscompiled by
gcc-4.9, but it is using a different register allocation ordering.

With the gcc-4.9 compiled function, something during the code patching
causes some of the %i* input registers to get corrupted.  Perhaps
we have a TLB miss path into the firmware that is deep enough to
cause a register window spill and subsequent restore when we get
back from the TLB miss trap.

Let's plug this up by doing two things:

1) Stop using the firmware stack for client interface calls into
   the firmware.  Just use the kernel's stack.

2) As soon as we can, call into a new function "start_early_boot()"
   to put a one-register-window buffer between the firmware's
   deepest stack frame and the top-most initial kernel one.

Reported-by: Meelis Roos <mroos@linux.ee>
Tested-by: Meelis Roos <mroos@linux.ee>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
David S. Miller
2014-10-23 12:58:13 -07:00
parent 61ed53deb1
commit ef3e035c3a
10 changed files with 40 additions and 62 deletions

View File

@@ -65,13 +65,10 @@ struct pause_patch_entry {
extern struct pause_patch_entry __pause_3insn_patch,
__pause_3insn_patch_end;
void __init per_cpu_patch(void);
void sun4v_patch_1insn_range(struct sun4v_1insn_patch_entry *,
struct sun4v_1insn_patch_entry *);
void sun4v_patch_2insn_range(struct sun4v_2insn_patch_entry *,
struct sun4v_2insn_patch_entry *);
void __init sun4v_patch(void);
void __init boot_cpu_id_too_large(int cpu);
extern unsigned int dcache_parity_tl1_occurred;
extern unsigned int icache_parity_tl1_occurred;

View File

@@ -672,14 +672,12 @@ tlb_fixup_done:
sethi %hi(init_thread_union), %g6
or %g6, %lo(init_thread_union), %g6
ldx [%g6 + TI_TASK], %g4
mov %sp, %l6
wr %g0, ASI_P, %asi
mov 1, %g1
sllx %g1, THREAD_SHIFT, %g1
sub %g1, (STACKFRAME_SZ + STACK_BIAS), %g1
add %g6, %g1, %sp
mov 0, %fp
/* Set per-cpu pointer initially to zero, this makes
* the boot-cpu use the in-kernel-image per-cpu areas
@@ -706,44 +704,14 @@ tlb_fixup_done:
nop
#endif
mov %l6, %o1 ! OpenPROM stack
call prom_init
mov %l7, %o0 ! OpenPROM cif handler
/* Initialize current_thread_info()->cpu as early as possible.
* In order to do that accurately we have to patch up the get_cpuid()
* assembler sequences. And that, in turn, requires that we know
* if we are on a Starfire box or not. While we're here, patch up
* the sun4v sequences as well.
/* To create a one-register-window buffer between the kernel's
* initial stack and the last stack frame we use from the firmware,
* do the rest of the boot from a C helper function.
*/
call check_if_starfire
nop
call per_cpu_patch
nop
call sun4v_patch
nop
#ifdef CONFIG_SMP
call hard_smp_processor_id
nop
cmp %o0, NR_CPUS
blu,pt %xcc, 1f
nop
call boot_cpu_id_too_large
nop
/* Not reached... */
1:
#else
mov 0, %o0
#endif
sth %o0, [%g6 + TI_CPU]
call prom_init_report
nop
/* Off we go.... */
call start_kernel
call start_early_boot
nop
/* Not reached... */

View File

@@ -109,7 +109,6 @@ hv_cpu_startup:
sllx %g5, THREAD_SHIFT, %g5
sub %g5, (STACKFRAME_SZ + STACK_BIAS), %g5
add %g6, %g5, %sp
mov 0, %fp
call init_irqwork_curcpu
nop

View File

@@ -30,6 +30,7 @@
#include <linux/cpu.h>
#include <linux/initrd.h>
#include <linux/module.h>
#include <linux/start_kernel.h>
#include <asm/io.h>
#include <asm/processor.h>
@@ -162,7 +163,7 @@ char reboot_command[COMMAND_LINE_SIZE];
static struct pt_regs fake_swapper_regs = { { 0, }, 0, 0, 0, 0 };
void __init per_cpu_patch(void)
static void __init per_cpu_patch(void)
{
struct cpuid_patch_entry *p;
unsigned long ver;
@@ -254,7 +255,7 @@ void sun4v_patch_2insn_range(struct sun4v_2insn_patch_entry *start,
}
}
void __init sun4v_patch(void)
static void __init sun4v_patch(void)
{
extern void sun4v_hvapi_init(void);
@@ -323,14 +324,25 @@ static void __init pause_patch(void)
}
}
#ifdef CONFIG_SMP
void __init boot_cpu_id_too_large(int cpu)
void __init start_early_boot(void)
{
prom_printf("Serious problem, boot cpu id (%d) >= NR_CPUS (%d)\n",
cpu, NR_CPUS);
prom_halt();
int cpu;
check_if_starfire();
per_cpu_patch();
sun4v_patch();
cpu = hard_smp_processor_id();
if (cpu >= NR_CPUS) {
prom_printf("Serious problem, boot cpu id (%d) >= NR_CPUS (%d)\n",
cpu, NR_CPUS);
prom_halt();
}
current_thread_info()->cpu = cpu;
prom_init_report();
start_kernel();
}
#endif
/* On Ultra, we support all of the v8 capabilities. */
unsigned long sparc64_elf_hwcap = (HWCAP_SPARC_FLUSH | HWCAP_SPARC_STBAR |

View File

@@ -109,10 +109,13 @@ startup_continue:
brnz,pn %g1, 1b
nop
sethi %hi(p1275buf), %g2
or %g2, %lo(p1275buf), %g2
ldx [%g2 + 0x10], %l2
add %l2, -(192 + 128), %sp
/* Get onto temporary stack which will be in the locked
* kernel image.
*/
sethi %hi(tramp_stack), %g1
or %g1, %lo(tramp_stack), %g1
add %g1, TRAMP_STACK_SIZE, %g1
sub %g1, STACKFRAME_SZ + STACK_BIAS + 256, %sp
flushw
/* Setup the loop variables:
@@ -394,7 +397,6 @@ after_lock_tlb:
sllx %g5, THREAD_SHIFT, %g5
sub %g5, (STACKFRAME_SZ + STACK_BIAS), %g5
add %g6, %g5, %sp
mov 0, %fp
rdpr %pstate, %o1
or %o1, PSTATE_IE, %o1