powerpc fixes for 5.5 #3

One fix for a regression introduced by our recent rework of cache flushing on
 memory hotunplug.
 
 Like several other arches, our VDSO clock_getres() needed a fix to match the
 semantics of posix_get_hrtimer_res().
 
 A fix for a boot crash on Power9 LPARs using PCI LSI interrupts.
 
 A commit disabling use of the trace_imc PMU (not the core PMU) on Power9
 systems, because it can lead to checkstops, until a workaround is developed.
 
 A handful of other minor fixes.
 
 Thanks to:
   Aneesh Kumar K.V, Anju T Sudhakar, Ard Biesheuvel, Christophe Leroy, Cédric Le
   Goater, Madhavan Srinivasan, Vincenzo Frascino.
 -----BEGIN PGP SIGNATURE-----
 
 iQJHBAABCAAxFiEEJFGtCPCthwEv2Y/bUevqPMjhpYAFAl3qRJcTHG1wZUBlbGxl
 cm1hbi5pZC5hdQAKCRBR6+o8yOGlgDymD/4lqWlFnXBUJA0MrYGa6PI7AA7W+oFC
 vVAFobQ0kCI7RiWv8UW1GKAOkQKIhIhCxl/TEv5DNevajelsvfaPeF/nlI5fL2y9
 klrvpMvQmEvBVTDispKQy3+pIr8EDWlvVOhFzU8YQMSZirxACLCmM/XNSmboCSR8
 Y3chFNgEL9nYpEVxX+IhN9oTJ8/eCJqfp8QhcVt1kG/jXnf8CVfSx7z/Qr1IarIT
 3LiiDBCCJbtG4WWH3Ku9WaCBZ66QxcTz/3RonGffYhalguhQjXl2vSdz9gg/icos
 G0H372ONZ5MkUj+IkyTTs/tmg2QVEzn1ZsVkstJZphbJ+/5VmrpdiAH0FvuJ/CaN
 kJ0dww4I9IEC2bl9Ok9XKNZSoQaXckfkPdX6JcHqD52pCIaBCoG+MdOvi0eMfcBj
 SDKuCK2BiAz+aOt0buSaqlgUNA6wXneb96h3I0u7CcV/CuzKChkJMLC6aJYB32DB
 gPZsnaespvOHBtbcLFP6RFAlY/nwaeLHTfSlLrbvs6k+5w9+ZUHA81uMzEBhNLN6
 ANPx57lDOzmlAQ/8fLyvGRWCdYZVGKmO4GaQXV9eDKn6lNV94hHkYMe0JX0bI0M9
 1SZP20t+zBBBV80nO7+JHEAxpTnT4RA6GNJ8p/bPyawEo6xdD1IMIZQy+EKYg++Q
 BeK76ccKp7wtZg==
 =1Q9O
 -----END PGP SIGNATURE-----

Merge tag 'powerpc-5.5-3' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux

Pull powerpc fixes from Michael Ellerman:
 "One fix for a regression introduced by our recent rework of cache
  flushing on memory hotunplug.

  Like several other arches, our VDSO clock_getres() needed a fix to
  match the semantics of posix_get_hrtimer_res().

  A fix for a boot crash on Power9 LPARs using PCI LSI interrupts.

  A commit disabling use of the trace_imc PMU (not the core PMU) on
  Power9 systems, because it can lead to checkstops, until a workaround
  is developed.

  A handful of other minor fixes.

  Thanks to: Aneesh Kumar K.V, Anju T Sudhakar, Ard Biesheuvel,
  Christophe Leroy, Cédric Le Goater, Madhavan Srinivasan, Vincenzo
  Frascino"

* tag 'powerpc-5.5-3' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/perf: Disable trace_imc pmu
  powerpc/powernv: Avoid re-registration of imc debugfs directory
  powerpc/pmem: Convert to EXPORT_SYMBOL_GPL
  powerpc/archrandom: fix arch_get_random_seed_int()
  powerpc: Fix vDSO clock_getres()
  powerpc/pmem: Fix kernel crash due to wrong range value usage in flush_dcache_range
  powerpc/xive: Skip ioremap() of ESB pages for LSI interrupts
  powerpc/kasan: Fix boot failure with RELOCATABLE && FSL_BOOKE
This commit is contained in:
Linus Torvalds 2019-12-06 13:34:31 -08:00
commit f89d416a86
11 changed files with 55 additions and 38 deletions

View File

@ -28,7 +28,7 @@ static inline int arch_get_random_seed_int(unsigned int *v)
unsigned long val;
int rc;
rc = arch_get_random_long(&val);
rc = arch_get_random_seed_long(&val);
if (rc)
*v = val;

View File

@ -83,6 +83,7 @@ struct vdso_data {
__s64 wtom_clock_sec; /* Wall to monotonic clock sec */
__s64 stamp_xtime_sec; /* xtime secs as at tb_orig_stamp */
__s64 stamp_xtime_nsec; /* xtime nsecs as at tb_orig_stamp */
__u32 hrtimer_res; /* hrtimer resolution */
__u32 syscall_map_64[SYSCALL_MAP_SIZE]; /* map of syscalls */
__u32 syscall_map_32[SYSCALL_MAP_SIZE]; /* map of syscalls */
};
@ -105,6 +106,7 @@ struct vdso_data {
__s32 stamp_xtime_sec; /* xtime seconds as at tb_orig_stamp */
__s32 stamp_xtime_nsec; /* xtime nsecs as at tb_orig_stamp */
__u32 stamp_sec_fraction; /* fractional seconds of stamp_xtime */
__u32 hrtimer_res; /* hrtimer resolution */
__u32 syscall_map_32[SYSCALL_MAP_SIZE]; /* map of syscalls */
__u32 dcache_block_size; /* L1 d-cache block size */
__u32 icache_block_size; /* L1 i-cache block size */

View File

@ -388,6 +388,7 @@ int main(void)
OFFSET(STAMP_XTIME_SEC, vdso_data, stamp_xtime_sec);
OFFSET(STAMP_XTIME_NSEC, vdso_data, stamp_xtime_nsec);
OFFSET(STAMP_SEC_FRAC, vdso_data, stamp_sec_fraction);
OFFSET(CLOCK_HRTIMER_RES, vdso_data, hrtimer_res);
OFFSET(CFG_ICACHE_BLOCKSZ, vdso_data, icache_block_size);
OFFSET(CFG_DCACHE_BLOCKSZ, vdso_data, dcache_block_size);
OFFSET(CFG_ICACHE_LOGBLOCKSZ, vdso_data, icache_log_block_size);
@ -413,7 +414,6 @@ int main(void)
DEFINE(CLOCK_REALTIME_COARSE, CLOCK_REALTIME_COARSE);
DEFINE(CLOCK_MONOTONIC_COARSE, CLOCK_MONOTONIC_COARSE);
DEFINE(NSEC_PER_SEC, NSEC_PER_SEC);
DEFINE(CLOCK_REALTIME_RES, MONOTONIC_RES_NSEC);
#ifdef CONFIG_BUG
DEFINE(BUG_ENTRY_SIZE, sizeof(struct bug_entry));

View File

@ -240,6 +240,9 @@ set_ivor:
bl early_init
#ifdef CONFIG_KASAN
bl kasan_early_init
#endif
#ifdef CONFIG_RELOCATABLE
mr r3,r30
mr r4,r31
@ -266,9 +269,6 @@ set_ivor:
/*
* Decide what sort of machine this is and initialize the MMU.
*/
#ifdef CONFIG_KASAN
bl kasan_early_init
#endif
mr r3,r30
mr r4,r31
bl machine_init

View File

@ -960,6 +960,7 @@ void update_vsyscall(struct timekeeper *tk)
vdso_data->stamp_xtime_sec = xt.tv_sec;
vdso_data->stamp_xtime_nsec = xt.tv_nsec;
vdso_data->stamp_sec_fraction = frac_sec;
vdso_data->hrtimer_res = hrtimer_resolution;
smp_wmb();
++(vdso_data->tb_update_count);
}

View File

@ -154,12 +154,15 @@ V_FUNCTION_BEGIN(__kernel_clock_getres)
cror cr0*4+eq,cr0*4+eq,cr1*4+eq
bne cr0,99f
mflr r12
.cfi_register lr,r12
bl __get_datapage@local /* get data page */
lwz r5, CLOCK_HRTIMER_RES(r3)
mtlr r12
li r3,0
cmpli cr0,r4,0
crclr cr0*4+so
beqlr
lis r5,CLOCK_REALTIME_RES@h
ori r5,r5,CLOCK_REALTIME_RES@l
stw r3,TSPC32_TV_SEC(r4)
stw r5,TSPC32_TV_NSEC(r4)
blr

View File

@ -186,12 +186,15 @@ V_FUNCTION_BEGIN(__kernel_clock_getres)
cror cr0*4+eq,cr0*4+eq,cr1*4+eq
bne cr0,99f
mflr r12
.cfi_register lr,r12
bl V_LOCAL_FUNC(__get_datapage)
lwz r5, CLOCK_HRTIMER_RES(r3)
mtlr r12
li r3,0
cmpldi cr0,r4,0
crclr cr0*4+so
beqlr
lis r5,CLOCK_REALTIME_RES@h
ori r5,r5,CLOCK_REALTIME_RES@l
std r3,TSPC64_TV_SEC(r4)
std r5,TSPC64_TV_NSEC(r4)
blr

View File

@ -17,14 +17,14 @@ void arch_wb_cache_pmem(void *addr, size_t size)
unsigned long start = (unsigned long) addr;
flush_dcache_range(start, start + size);
}
EXPORT_SYMBOL(arch_wb_cache_pmem);
EXPORT_SYMBOL_GPL(arch_wb_cache_pmem);
void arch_invalidate_pmem(void *addr, size_t size)
{
unsigned long start = (unsigned long) addr;
flush_dcache_range(start, start + size);
}
EXPORT_SYMBOL(arch_invalidate_pmem);
EXPORT_SYMBOL_GPL(arch_invalidate_pmem);
/*
* CONFIG_ARCH_HAS_UACCESS_FLUSHCACHE symbols

View File

@ -121,7 +121,7 @@ static void flush_dcache_range_chunked(unsigned long start, unsigned long stop,
unsigned long i;
for (i = start; i < stop; i += chunk) {
flush_dcache_range(i, min(stop, start + chunk));
flush_dcache_range(i, min(stop, i + chunk));
cond_resched();
}
}

View File

@ -59,10 +59,6 @@ static void export_imc_mode_and_cmd(struct device_node *node,
imc_debugfs_parent = debugfs_create_dir("imc", powerpc_debugfs_root);
/*
* Return here, either because 'imc' directory already exists,
* Or failed to create a new one.
*/
if (!imc_debugfs_parent)
return;
@ -135,7 +131,6 @@ static int imc_get_mem_addr_nest(struct device_node *node,
}
pmu_ptr->imc_counter_mmaped = true;
export_imc_mode_and_cmd(node, pmu_ptr);
kfree(base_addr_arr);
kfree(chipid_arr);
return 0;
@ -151,7 +146,7 @@ error:
* and domain as the inputs.
* Allocates memory for the struct imc_pmu, sets up its domain, size and offsets
*/
static int imc_pmu_create(struct device_node *parent, int pmu_index, int domain)
static struct imc_pmu *imc_pmu_create(struct device_node *parent, int pmu_index, int domain)
{
int ret = 0;
struct imc_pmu *pmu_ptr;
@ -159,27 +154,23 @@ static int imc_pmu_create(struct device_node *parent, int pmu_index, int domain)
/* Return for unknown domain */
if (domain < 0)
return -EINVAL;
return NULL;
/* memory for pmu */
pmu_ptr = kzalloc(sizeof(*pmu_ptr), GFP_KERNEL);
if (!pmu_ptr)
return -ENOMEM;
return NULL;
/* Set the domain */
pmu_ptr->domain = domain;
ret = of_property_read_u32(parent, "size", &pmu_ptr->counter_mem_size);
if (ret) {
ret = -EINVAL;
if (ret)
goto free_pmu;
}
if (!of_property_read_u32(parent, "offset", &offset)) {
if (imc_get_mem_addr_nest(parent, pmu_ptr, offset)) {
ret = -EINVAL;
if (imc_get_mem_addr_nest(parent, pmu_ptr, offset))
goto free_pmu;
}
}
/* Function to register IMC pmu */
@ -190,14 +181,14 @@ static int imc_pmu_create(struct device_node *parent, int pmu_index, int domain)
if (pmu_ptr->domain == IMC_DOMAIN_NEST)
kfree(pmu_ptr->mem_info);
kfree(pmu_ptr);
return ret;
return NULL;
}
return 0;
return pmu_ptr;
free_pmu:
kfree(pmu_ptr);
return ret;
return NULL;
}
static void disable_nest_pmu_counters(void)
@ -254,6 +245,7 @@ int get_max_nest_dev(void)
static int opal_imc_counters_probe(struct platform_device *pdev)
{
struct device_node *imc_dev = pdev->dev.of_node;
struct imc_pmu *pmu;
int pmu_count = 0, domain;
bool core_imc_reg = false, thread_imc_reg = false;
u32 type;
@ -269,6 +261,7 @@ static int opal_imc_counters_probe(struct platform_device *pdev)
}
for_each_compatible_node(imc_dev, NULL, IMC_DTB_UNIT_COMPAT) {
pmu = NULL;
if (of_property_read_u32(imc_dev, "type", &type)) {
pr_warn("IMC Device without type property\n");
continue;
@ -285,7 +278,14 @@ static int opal_imc_counters_probe(struct platform_device *pdev)
domain = IMC_DOMAIN_THREAD;
break;
case IMC_TYPE_TRACE:
domain = IMC_DOMAIN_TRACE;
/*
* FIXME. Using trace_imc events to monitor application
* or KVM thread performance can cause a checkstop
* (system crash).
* Disable it for now.
*/
pr_info_once("IMC: disabling trace_imc PMU\n");
domain = -1;
break;
default:
pr_warn("IMC Unknown Device type \n");
@ -293,9 +293,13 @@ static int opal_imc_counters_probe(struct platform_device *pdev)
break;
}
if (!imc_pmu_create(imc_dev, pmu_count, domain)) {
if (domain == IMC_DOMAIN_NEST)
pmu = imc_pmu_create(imc_dev, pmu_count, domain);
if (pmu != NULL) {
if (domain == IMC_DOMAIN_NEST) {
if (!imc_debugfs_parent)
export_imc_mode_and_cmd(imc_dev, pmu);
pmu_count++;
}
if (domain == IMC_DOMAIN_CORE)
core_imc_reg = true;
if (domain == IMC_DOMAIN_THREAD)
@ -303,10 +307,6 @@ static int opal_imc_counters_probe(struct platform_device *pdev)
}
}
/* If none of the nest units are registered, remove debugfs interface */
if (pmu_count == 0)
debugfs_remove_recursive(imc_debugfs_parent);
/* If core imc is not registered, unregister thread-imc */
if (!core_imc_reg && thread_imc_reg)
unregister_thread_imc();

View File

@ -392,20 +392,28 @@ static int xive_spapr_populate_irq_data(u32 hw_irq, struct xive_irq_data *data)
data->esb_shift = esb_shift;
data->trig_page = trig_page;
data->hw_irq = hw_irq;
/*
* No chip-id for the sPAPR backend. This has an impact how we
* pick a target. See xive_pick_irq_target().
*/
data->src_chip = XIVE_INVALID_CHIP_ID;
/*
* When the H_INT_ESB flag is set, the H_INT_ESB hcall should
* be used for interrupt management. Skip the remapping of the
* ESB pages which are not available.
*/
if (data->flags & XIVE_IRQ_FLAG_H_INT_ESB)
return 0;
data->eoi_mmio = ioremap(data->eoi_page, 1u << data->esb_shift);
if (!data->eoi_mmio) {
pr_err("Failed to map EOI page for irq 0x%x\n", hw_irq);
return -ENOMEM;
}
data->hw_irq = hw_irq;
/* Full function page supports trigger */
if (flags & XIVE_SRC_TRIGGER) {
data->trig_mmio = data->eoi_mmio;