process/vm; replace vm_range list by a rbtree

This replaces the chained list used to keep track of all memory ranges
of a process by a standard rbtree (no need of interval tree here
because there is no overlap)

Accesses that were done directly through vm_range_list before were
replaced by lookup_process_memory_range, even full list scan (e.g.
coredump).
The full scans will thus be less efficient because calls to rb_next()
will not be inlined, but these are rarer calls that can probably afford
this compared to code simplicity.

The only reference to the actual backing structure left outside of
process.c is a call to rb_erase in xpmem_free_process_memory_range.

v2: fix lookup_process_memory_range with small start address

v3: make vm_range_insert error out properly

Panic does not lead to easy debug, all error paths
are handled to just return someting on error

v4: fix lookup_process_memory_range (again)

That optimistically going left was a more serious bug than just
last iteration, we could just pass by a match and continue down
the tree if the match was not a leaf.

v5: some users actually needed leftmost match, so restore behavior
without the breakage (hopefully)

arch/arm64/kernel/cpu.c

@@ -1499,7 +1499,6 @@ unhandled_page_fault(struct thread *thread, void *fault_addr, void *regs)
 	const uintptr_t address = (uintptr_t)fault_addr;
 	struct process_vm *vm = thread->vm;
 	struct vm_range *range;
-	char found;
 	unsigned long irqflags;
 	unsigned long error = 0;
 
@@ -1513,17 +1512,12 @@ unhandled_page_fault(struct thread *thread, void *fault_addr, void *regs)
 			(error & PF_RSVD ? "was" : "wasn't"),
 			(error & PF_INSTR ? "was" : "wasn't"));
 
-	found = 0;
-	list_for_each_entry(range, &vm->vm_range_list, list) {
-		if (range->start <= address && range->end > address) {
-			found = 1;
-			__kprintf("address is in range, flag: 0x%lx\n",
-					range->flag);
-			ihk_mc_pt_print_pte(vm->address_space->page_table, (void*)address);
-			break;
-		}
-	}
-	if (!found) {
+	range = lookup_process_memory_range(vm, address, address+1);
+	if (range) {
+		__kprintf("address is in range, flag: 0x%lx\n",
+				range->flag);
+		ihk_mc_pt_print_pte(vm->address_space->page_table, (void*)address);
+	} else {
 		__kprintf("address is out of range! \n");
 	}
 

arch/arm64/kernel/gencore.c

@@ -221,7 +221,7 @@ int gencore(struct thread *thread, void *regs,
 	Elf64_Ehdr eh;
 	Elf64_Phdr *ph = NULL;
 	void *note = NULL;
-	struct vm_range *range;
+	struct vm_range *range, *next;
 	struct process_vm *vm = thread->vm;
 	int segs = 1;	/* the first one is for NOTE */
 	int notesize, phsize, alignednotesize;
@@ -235,7 +235,10 @@ int gencore(struct thread *thread, void *regs,
 		return -1;
 	}
 
-	list_for_each_entry(range, &vm->vm_range_list, list) {
+	next = lookup_process_memory_range(vm, 0, -1);
+	while ((range = next)) {
+		next = next_process_memory_range(vm, range);
+
 		dkprintf("start:%lx end:%lx flag:%lx objoff:%lx\n", 
 			 range->start, range->end, range->flag, range->objoff);
 		/* We omit reserved areas because they are only for
@@ -323,7 +326,10 @@ int gencore(struct thread *thread, void *regs,
 
 	/* program header for each memory chunk */
 	i = 1;
-	list_for_each_entry(range, &vm->vm_range_list, list) {
+	next = lookup_process_memory_range(vm, 0, -1);
+	while ((range = next)) {
+		next = next_process_memory_range(vm, range);
+
 		unsigned long flag = range->flag;
 		unsigned long size = range->end - range->start;
 
@@ -364,7 +370,10 @@ int gencore(struct thread *thread, void *regs,
 	dkprintf("coretable[2]: %lx@%lx(%lx)\n", ct[2].len, ct[2].addr, note);
 
 	i = 3;	/* memory segments */
-	list_for_each_entry(range, &vm->vm_range_list, list) {
+	next = lookup_process_memory_range(vm, 0, -1);
+	while ((range = next)) {
+		next = next_process_memory_range(vm, range);
+
 		unsigned long phys;
 
 		if (range->flag & VR_RESERVED)

arch/x86/kernel/cpu.c

@@ -1081,7 +1081,6 @@ unhandled_page_fault(struct thread *thread, void *fault_addr, void *regs)
 	const uintptr_t address = (uintptr_t)fault_addr;
 	struct process_vm *vm = thread->vm;
 	struct vm_range *range;
-	char found;
 	unsigned long irqflags;
 	unsigned long error = ((struct x86_user_context *)regs)->gpr.error;
 
@@ -1095,17 +1094,12 @@ unhandled_page_fault(struct thread *thread, void *fault_addr, void *regs)
 			(error & PF_RSVD ? "was" : "wasn't"),
 			(error & PF_INSTR ? "was" : "wasn't"));
 
-	found = 0;
-	list_for_each_entry(range, &vm->vm_range_list, list) {
-		if (range->start <= address && range->end > address) {
-			found = 1;
-			__kprintf("address is in range, flag: 0x%lx\n",
-					range->flag);
-			ihk_mc_pt_print_pte(vm->address_space->page_table, (void*)address);
-			break;
-		}
-	}
-	if (!found) {
+	range = lookup_process_memory_range(vm, address, address+1);
+	if (range) {
+		__kprintf("address is in range, flag: 0x%lx\n",
+				range->flag);
+		ihk_mc_pt_print_pte(vm->address_space->page_table, (void*)address);
+	} else {
 		__kprintf("address is out of range! \n");
 	}
 

arch/x86/kernel/gencore.c

@@ -289,7 +289,7 @@ int gencore(struct thread *thread, void *regs,
 	Elf64_Ehdr eh;
 	Elf64_Phdr *ph = NULL;
 	void *note = NULL;
-	struct vm_range *range;
+	struct vm_range *range, *next;
 	struct process_vm *vm = thread->vm;
 	int segs = 1;	/* the first one is for NOTE */
 	int notesize, phsize, alignednotesize;
@@ -303,7 +303,10 @@ int gencore(struct thread *thread, void *regs,
 		return -1;
 	}
 
-	list_for_each_entry(range, &vm->vm_range_list, list) {
+	next = lookup_process_memory_range(vm, 0, -1);
+	while ((range = next)) {
+		next = next_process_memory_range(vm, range);
+
 		dkprintf("start:%lx end:%lx flag:%lx objoff:%lx\n", 
 			 range->start, range->end, range->flag, range->objoff);
 		/* We omit reserved areas because they are only for
@@ -393,7 +396,10 @@ int gencore(struct thread *thread, void *regs,
 
 	/* program header for each memory chunk */
 	i = 1;
-	list_for_each_entry(range, &vm->vm_range_list, list) {
+	next = lookup_process_memory_range(vm, 0, -1);
+	while ((range = next)) {
+		next = next_process_memory_range(vm, range);
+
 		unsigned long flag = range->flag;
 		unsigned long size = range->end - range->start;
 
@@ -434,7 +440,10 @@ int gencore(struct thread *thread, void *regs,
 	dkprintf("coretable[2]: %lx@%lx(%lx)\n", ct[2].len, ct[2].addr, note);
 
 	i = 3;	/* memory segments */
-	list_for_each_entry(range, &vm->vm_range_list, list) {
+	next = lookup_process_memory_range(vm, 0, -1);
+	while ((range = next)) {
+		next = next_process_memory_range(vm, range);
+
 		unsigned long phys;
 
 		if (range->flag & VR_RESERVED)