| Message ID | 1452867647-1410-1-git-send-email-grygorii.strashko@ti.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
* Grygorii Strashko | 2016-01-15 16:20:47 [+0200]: >Now the random crashes are observed on ARM TI am437x-idk-evm used with >-RT Kernel v4.4-rt2 and with Highmem enabled. > >The reason of issue is that, newly introduced switch_kmaps() does not >take into account FIX_KMAP_BEGIN, which was re-added by >commit a5f4c561b3b1 ("ARM: 8415/1: early fixmap support for earlycon") >(K4.3) for ARM and since that the value of FIX_KMAP_BEGIN >is not 0 any more. > >Hence update switch_kmaps() so it will take into FIX_KMAP_BEGIN while >calculating fixmap idx. This patch also introduces fixmap_idx() to >make code simpler and remove duplicated code. Thank you debugging and fixing this. I will fold this into the initial patch since it is a v4.1..v4.4 fallout. Regarding your "nobody else be me" question in the other mail: once I managed something ARM+highmem I I see this: |Welcome to Debian GNU/Linux stretch/sid! | |[ 11.738637] systemd[1]: Set hostname to <vexpress>. |[ 11.909949] ------------[ cut here ]------------ |[ 11.909958] kernel BUG at arch/arm/mm/highmem.c:90! So I did not get as far as a shell to start anything. >Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> >Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> Sebastian
* Grygorii Strashko | 2016-01-15 16:20:47 [+0200]: >--- a/arch/arm/mm/highmem.c >+++ b/arch/arm/mm/highmem.c >@@ -175,7 +180,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) > * Restore @next_p's kmap_atomic mappings > */ > for (i = 0; i < next_p->kmap_idx; i++) { >- int idx = i + KM_TYPE_NR * smp_processor_id(); >+ int idx = fixmap_idx(i); > > if (!pte_none(next_p->kmap_pte[i])) > set_fixmap_pte(idx, next_p->kmap_pte[i]); Grygorii, if you remove this chunk then it should apply upstream. Could you forward it upstream then please? So the -RT piece would shrink a little:) Sebastian
On Wed, Jan 20, 2016 at 09:21:39PM +0100, Sebastian Andrzej Siewior wrote: > * Grygorii Strashko | 2016-01-15 16:20:47 [+0200]: > > >--- a/arch/arm/mm/highmem.c > >+++ b/arch/arm/mm/highmem.c > >@@ -175,7 +180,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) > > * Restore @next_p's kmap_atomic mappings > > */ > > for (i = 0; i < next_p->kmap_idx; i++) { > >- int idx = i + KM_TYPE_NR * smp_processor_id(); > >+ int idx = fixmap_idx(i); > > > > if (!pte_none(next_p->kmap_pte[i])) > > set_fixmap_pte(idx, next_p->kmap_pte[i]); > > Grygorii, if you remove this chunk then it should apply upstream. Could > you forward it upstream then please? So the -RT piece would shrink a > little:) Or I could commit my own patch which forms a proportion of this change.
On 01/20/2016 10:23 PM, Russell King - ARM Linux wrote: > On Wed, Jan 20, 2016 at 09:21:39PM +0100, Sebastian Andrzej Siewior wrote: >> * Grygorii Strashko | 2016-01-15 16:20:47 [+0200]: >> >>> --- a/arch/arm/mm/highmem.c >>> +++ b/arch/arm/mm/highmem.c >>> @@ -175,7 +180,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) >>> * Restore @next_p's kmap_atomic mappings >>> */ >>> for (i = 0; i < next_p->kmap_idx; i++) { >>> - int idx = i + KM_TYPE_NR * smp_processor_id(); >>> + int idx = fixmap_idx(i); >>> >>> if (!pte_none(next_p->kmap_pte[i])) >>> set_fixmap_pte(idx, next_p->kmap_pte[i]); >> >> Grygorii, if you remove this chunk then it should apply upstream. Could >> you forward it upstream then please? So the -RT piece would shrink a >> little:) > > Or I could commit my own patch which forms a proportion of this change. > np from my side.
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c index deabc36..542692d 100644 --- a/arch/arm/mm/highmem.c +++ b/arch/arm/mm/highmem.c @@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr) return *ptep; } +static unsigned int fixmap_idx(int type) +{ + return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); +} + void *kmap(struct page *page) { might_sleep(); @@ -80,7 +85,7 @@ void *kmap_atomic(struct page *page) type = kmap_atomic_idx_push(); - idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); + idx = fixmap_idx(type); vaddr = __fix_to_virt(idx); #ifdef CONFIG_DEBUG_HIGHMEM /* @@ -110,7 +115,7 @@ void __kunmap_atomic(void *kvaddr) if (kvaddr >= (void *)FIXADDR_START) { type = kmap_atomic_idx(); - idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); + idx = fixmap_idx(type); if (cache_is_vivt()) __cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE); @@ -146,7 +151,7 @@ void *kmap_atomic_pfn(unsigned long pfn) return page_address(page); type = kmap_atomic_idx_push(); - idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); + idx = fixmap_idx(type); vaddr = __fix_to_virt(idx); #ifdef CONFIG_DEBUG_HIGHMEM BUG_ON(!pte_none(get_fixmap_pte(vaddr))); @@ -167,7 +172,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) * Clear @prev's kmap_atomic mappings */ for (i = 0; i < prev_p->kmap_idx; i++) { - int idx = i + KM_TYPE_NR * smp_processor_id(); + int idx = fixmap_idx(i); set_fixmap_pte(idx, __pte(0)); } @@ -175,7 +180,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) * Restore @next_p's kmap_atomic mappings */ for (i = 0; i < next_p->kmap_idx; i++) { - int idx = i + KM_TYPE_NR * smp_processor_id(); + int idx = fixmap_idx(i); if (!pte_none(next_p->kmap_pte[i])) set_fixmap_pte(idx, next_p->kmap_pte[i]);
Now the random crashes are observed on ARM TI am437x-idk-evm used with -RT Kernel v4.4-rt2 and with Highmem enabled. The reason of issue is that, newly introduced switch_kmaps() does not take into account FIX_KMAP_BEGIN, which was re-added by commit a5f4c561b3b1 ("ARM: 8415/1: early fixmap support for earlycon") (K4.3) for ARM and since that the value of FIX_KMAP_BEGIN is not 0 any more. Hence update switch_kmaps() so it will take into FIX_KMAP_BEGIN while calculating fixmap idx. This patch also introduces fixmap_idx() to make code simpler and remove duplicated code. Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> --- arch/arm/mm/highmem.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)