We had been doing an automatic full eviction of the L1 I$
everywhere whenever we did a kernel-space TLB flush. It turns
out this isn't necessary, since all the callers already handle
doing a flush if necessary.
Signed-off-by: Chris Metcalf <[email protected]>
---
arch/tile/kernel/tlb.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/arch/tile/kernel/tlb.c b/arch/tile/kernel/tlb.c
index 3fd54d5..f23b535 100644
--- a/arch/tile/kernel/tlb.c
+++ b/arch/tile/kernel/tlb.c
@@ -91,8 +91,14 @@ void flush_tlb_all(void)
}
}
+/*
+ * Callers need to flush the L1I themselves if necessary, e.g. for
+ * kernel module unload. Otherwise we assume callers are not using
+ * executable pgprot_t's. Using EVICT_L1I means that dataplane cpus
+ * will get an unnecessary interrupt otherwise.
+ */
void flush_tlb_kernel_range(unsigned long start, unsigned long end)
{
- flush_remote(0, HV_FLUSH_EVICT_L1I, cpu_online_mask,
+ flush_remote(0, 0, NULL,
start, end - start, PAGE_SIZE, cpu_online_mask, NULL, 0);
}
--
1.8.3.1