linux-kernel - Re: can't flush tlb on e500

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20090522.095720.68565257.saito@densan.co.jp>
Date:	Fri, 22 May 2009 09:57:20 +0900 (JST)
From:	Hideo Saito <hsaito.ppc@...il.com>
To:	linux-kernel@...r.kernel.org
Cc:	benh@...nel.crashing.org, hsaito.ppc@...il.com
Subject: Re: can't flush tlb on e500

How about following changes because all TLB entries are flushed repeatedly if processes overflow at the table mapping the context. I think that the table should be initialized again because _tlbil_pid() flushes all TLB entries and tlbilx instruction isn't supported on E500(MPC8548).

--- arch/powerpc/mm/mmu_context_nohash.c.orig	2009-03-24 08:12:14.000000000 +0900
+++ arch/powerpc/mm/mmu_context_nohash.c	2009-05-21 16:35:09.000000000 +0900
@@ -107,39 +107,69 @@ static unsigned int steal_context_smp(un
 	 */
 	spin_unlock(&context_lock);
 	cpu_relax();
 	spin_lock(&context_lock);
 	goto again;
 }
 #endif  /* CONFIG_SMP */
 
+/*
+ * We're flushed using the all context
+ */
+static void flush_all_context(int cpu)
+{
+	struct mm_struct *mm;
+	int n;
+
+	for (n = first_context; n <= last_context; n++) {
+
+		mm = context_mm[n];
+		if (mm == NULL || mm->context.id == MMU_NO_CONTEXT)
+			continue;
+
+		WARN_ON(mm->context.active != 0);
+
+		mm->context.id = MMU_NO_CONTEXT;
+	}
+	memset(stale_map[cpu], 0, CTX_MAP_SIZE);
+	memset(context_map, 0, CTX_MAP_SIZE);
+	context_map[0] = (1 << first_context) - 1;
+	nr_free_contexts = last_context - first_context + 1;
+}
+
 /* Note that this will also be called on SMP if all other CPUs are
  * offlined, which means that it may be called for cpu != 0. For
  * this to work, we somewhat assume that CPUs that are onlined
  * come up with a fully clean TLB (or are cleaned when offlined)
  */
 static unsigned int steal_context_up(unsigned int id)
 {
 	struct mm_struct *mm;
 	int cpu = smp_processor_id();
 
 	/* Pick up the victim mm */
 	mm = context_mm[id];
 
 	pr_debug("[%d] steal context %d from mm @%p\n", cpu, id, mm);
 
-	/* Mark this mm has having no context anymore */
-	mm->context.id = MMU_NO_CONTEXT;
-
 	/* Flush the TLB for that context */
 	local_flush_tlb_mm(mm);
 
+#ifdef CONFIG_FSL_BOOKE
+	flush_all_context(cpu);
+	__set_bit(id, context_map);
+	nr_free_contexts--;
+#else
+	/* Mark this mm has having no context anymore */
+	mm->context.id = MMU_NO_CONTEXT;
+
 	/* XXX This clear should ultimately be part of local_flush_tlb_mm */
 	__clear_bit(id, stale_map[cpu]);
+#endif
 
 	return id;
 }
 
 #ifdef DEBUG_MAP_CONSISTENCY
 static void context_check_map(void)
 {
 	unsigned int id, nrf, nact;
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/