Skip to contentSkip to conversation

Fix recent filter regressions

commit for review Mathieu Desnoyers 2 participants 2 messages Sep 10, 2012 2 messages not in the archive

lore
Share & tools
lore
#1 lore raw
In reply to a message not in the archive
Quoted 4 lines · * Mathieu Desnoyers ([email protected]) wrote:

Hi!

Please review this patch from my pull queue for lttng-tools. As soon as we get your Acked-by, David will pull it.

As per our private irc discussion, how about the following update ?

commit aa06477c0d844bb5b68c8f8070fd0cae3a80de52

Author: Mathieu Desnoyers <[email protected]>
Date:   Mon Sep 10 10:45:41 2012 -0400

Fix filter: fix recent regressions

Fix:
commit d93c4f1ffcffa73102e3299276f2f83951a68c36
Author: Christian Babeux <[email protected]>
Date: Thu Sep 6 13:40:19 2012 -0400

Fix: Accept bytecode of length 65536 bytes

Broke the filter: it changed the reloc table format, without knowledge of UST, and without good reason: it should stay { uint16_t, string } -- it does not need a uint32_t offset. Add a check for the value returned by bytecode_get_len() when populating the reloc table.

Fix:
commit 5ddb0a08c5c3ca917b025032b864d78e53c7c68a
Author: Christian Babeux <[email protected]>
Date: Mon Aug 27 14:48:21 2012 -0400

Fix: Generation of bytecode longer than 32768 bytes fails

Mixed the concepts of "len" and "alloc_len". It was therefore not checking the bounds correctly, and was not zeroing the memory range expected to be zeroed, causing out-of-bound array access.

src/lib/lttng-ctl/filter/filter-visitor-generate-bytecode.c +18 -12
index 332a387..8d44f4b 100644
@@ -94,21 +94,21 @@ int32_t bytecode_reserve(struct lttng_filter_bytecode_alloc **fb, uint32_t align
 {
 	int32_t ret;
 	uint32_t padding = offset_align((*fb)->b.len, align);
+	uint32_t new_len = (*fb)->b.len + padding + len;
+	uint32_t new_alloc_len = sizeof(struct lttng_filter_bytecode) + new_len;
+	uint32_t old_alloc_len = (*fb)->alloc_len;
 
-	if ((*fb)->b.len + padding + len > LTTNG_FILTER_MAX_LEN)
+	if (new_len > LTTNG_FILTER_MAX_LEN)
 		return -EINVAL;
 
-	if ((*fb)->b.len + padding + len > (*fb)->alloc_len) {
-		uint32_t new_len =
-			max_t(uint32_t, 1U << get_count_order((*fb)->b.len + padding + len),
-				(*fb)->alloc_len << 1);
-		uint32_t old_len = (*fb)->alloc_len;
-
-		*fb = realloc(*fb, sizeof(struct lttng_filter_bytecode_alloc) + new_len);
+	if (new_alloc_len > old_alloc_len) {
+		new_alloc_len =
+			max_t(uint32_t, 1U << get_count_order(new_alloc_len), old_alloc_len << 1);
+		*fb = realloc(*fb, new_alloc_len);
 		if (!*fb)
 			return -ENOMEM;
-		memset(&(*fb)->b.data[old_len], 0, new_len - old_len);
-		(*fb)->alloc_len = new_len;
+		memset(&((char *) *fb)[old_alloc_len], 0, new_alloc_len - old_alloc_len);
+		(*fb)->alloc_len = new_alloc_len;
 	}
 	(*fb)->b.len += padding;
 	ret = (*fb)->b.len;
@@ -239,7 +239,8 @@ int visit_node_load(struct filter_parser_ctx *ctx, struct ir_op *node)
 		uint32_t insn_len = sizeof(struct load_op)
 			+ sizeof(struct field_ref);
 		struct field_ref ref_offset;
-		uint32_t reloc_offset;
+		uint32_t reloc_offset_u32;
+		uint16_t reloc_offset;
 
 		insn = calloc(insn_len, 1);
 		if (!insn)
@@ -248,7 +249,12 @@ int visit_node_load(struct filter_parser_ctx *ctx, struct ir_op *node)
 		ref_offset.offset = (uint16_t) -1U;
 		memcpy(insn->data, &ref_offset, sizeof(ref_offset));
 		/* reloc_offset points to struct load_op */
-		reloc_offset = bytecode_get_len(&ctx->bytecode->b);
+		reloc_offset_u32 = bytecode_get_len(&ctx->bytecode->b);
+		if (reloc_offset_u32 > LTTNG_FILTER_MAX_LEN - 1) {
+			free(insn);
+			return -EINVAL;
+		}
+		reloc_offset = (uint16_t) reloc_offset_u32;
 		ret = bytecode_push(&ctx->bytecode, insn, 1, insn_len);
 		if (ret) {
 			free(insn);
Signature · 4 lines
Mathieu Desnoyers
Operating System Efficiency R&D Consultant
EfficiOS Inc.
http://www.efficios.com
#3 lore raw
In reply to a message not in the archive

Looks good to me!

Quoted 105 lines · Mathieu Desnoyers
Quoted 4 lines · * Mathieu Desnoyers ([email protected]) wrote:

Hi!

Please review this patch from my pull queue for lttng-tools. As soon as we get your Acked-by, David will pull it.

As per our private irc discussion, how about the following update ?

commit aa06477c0d844bb5b68c8f8070fd0cae3a80de52

Author: Mathieu Desnoyers <[email protected]>
Date:   Mon Sep 10 10:45:41 2012 -0400

Fix filter: fix recent regressions

Fix:
commit d93c4f1ffcffa73102e3299276f2f83951a68c36
Author: Christian Babeux <[email protected]>
Date: Thu Sep 6 13:40:19 2012 -0400

Fix: Accept bytecode of length 65536 bytes

Broke the filter: it changed the reloc table format, without knowledge of UST, and without good reason: it should stay { uint16_t, string } -- it does not need a uint32_t offset. Add a check for the value returned by bytecode_get_len() when populating the reloc table.

Fix:
commit 5ddb0a08c5c3ca917b025032b864d78e53c7c68a
Author: Christian Babeux <[email protected]>
Date: Mon Aug 27 14:48:21 2012 -0400

Fix: Generation of bytecode longer than 32768 bytes fails

Mixed the concepts of "len" and "alloc_len". It was therefore not checking the bounds correctly, and was not zeroing the memory range expected to be zeroed, causing out-of-bound array access.

src/lib/lttng-ctl/filter/filter-visitor-generate-bytecode.c +18 -13
index 332a387..8d44f4b 100644
@@ -94,21 +94,21 @@ int32_t bytecode_reserve(struct lttng_filter_bytecode_alloc **fb, uint32_t align
 {
        int32_t ret;
        uint32_t padding = offset_align((*fb)->b.len, align);
+       uint32_t new_len = (*fb)->b.len + padding + len;
+       uint32_t new_alloc_len = sizeof(struct lttng_filter_bytecode) + new_len;
+       uint32_t old_alloc_len = (*fb)->alloc_len;
 
-       if ((*fb)->b.len + padding + len > LTTNG_FILTER_MAX_LEN)
+       if (new_len > LTTNG_FILTER_MAX_LEN)
                return -EINVAL;
 
-       if ((*fb)->b.len + padding + len > (*fb)->alloc_len) {
-               uint32_t new_len =
-                       max_t(uint32_t, 1U << get_count_order((*fb)->b.len + padding + len),
-                               (*fb)->alloc_len << 1);
-               uint32_t old_len = (*fb)->alloc_len;
-
-               *fb = realloc(*fb, sizeof(struct lttng_filter_bytecode_alloc) + new_len);
+       if (new_alloc_len > old_alloc_len) {
+               new_alloc_len =
+                       max_t(uint32_t, 1U << get_count_order(new_alloc_len), old_alloc_len << 1);
+               *fb = realloc(*fb, new_alloc_len);
                if (!*fb)
                        return -ENOMEM;
-               memset(&(*fb)->b.data[old_len], 0, new_len - old_len);
-               (*fb)->alloc_len = new_len;
+               memset(&((char *) *fb)[old_alloc_len], 0, new_alloc_len - old_alloc_len);
+               (*fb)->alloc_len = new_alloc_len;
        }
        (*fb)->b.len += padding;
        ret = (*fb)->b.len;
@@ -239,7 +239,8 @@ int visit_node_load(struct filter_parser_ctx *ctx, struct ir_op *node)
                uint32_t insn_len = sizeof(struct load_op)
                        + sizeof(struct field_ref);
                struct field_ref ref_offset;
-               uint32_t reloc_offset;
+               uint32_t reloc_offset_u32;
+               uint16_t reloc_offset;
 
                insn = calloc(insn_len, 1);
                if (!insn)
@@ -248,7 +249,12 @@ int visit_node_load(struct filter_parser_ctx *ctx, struct ir_op *node)
                ref_offset.offset = (uint16_t) -1U;
                memcpy(insn->data, &ref_offset, sizeof(ref_offset));
                /* reloc_offset points to struct load_op */
-               reloc_offset = bytecode_get_len(&ctx->bytecode->b);
+               reloc_offset_u32 = bytecode_get_len(&ctx->bytecode->b);
+               if (reloc_offset_u32 > LTTNG_FILTER_MAX_LEN - 1) {
+                       free(insn);
+                       return -EINVAL;
+               }
+               reloc_offset = (uint16_t) reloc_offset_u32;
                ret = bytecode_push(&ctx->bytecode, insn, 1, insn_len);
                if (ret) {
                        free(insn);
 
--

Mathieu Desnoyers
Operating System Efficiency R&D Consultant
EfficiOS Inc.
http://www.efficios.com