This is an automated email from the ASF dual-hosted git repository.
xiaoxiang pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/nuttx.git
commit 3f3bbf76f94bc858784140127f6ce3b52f70bf61
Author: buxiasen <buxia...@xiaomi.com>
AuthorDate: Wed Aug 28 19:56:36 2024 +0800
tool/gdb/memdump: fix orphan, fix -p support
Signed-off-by: buxiasen <buxia...@xiaomi.com>
---
tools/gdb/memdump.py | 75 ++++++++++++++++++++++++++++------------------------
1 file changed, 40 insertions(+), 35 deletions(-)
diff --git a/tools/gdb/memdump.py b/tools/gdb/memdump.py
index ff09c36b15..cee921ecca 100644
--- a/tools/gdb/memdump.py
+++ b/tools/gdb/memdump.py
@@ -295,14 +295,14 @@ class HeapNode:
record_size = gdb_node["size"]
- if hasattr(gdb_node, "seqno"):
+ try:
seqno = gdb_node["seqno"]
- else:
+ except gdb.error:
seqno = 0
- if hasattr(gdb_node, "pid"):
+ try:
node_pid = gdb_node["pid"]
- else:
+ except gdb.error:
node_pid = 0
self.size = mm_nodesize(record_size)
@@ -424,6 +424,34 @@ class Memdump(gdb.Command):
backtrace_dict=self.backtrace_dict,
)
+ def memdump_tail(self, detail, simple):
+ if not detail:
+ output = [v for v in self.backtrace_dict.values()]
+ output.sort(key=get_count, reverse=True)
+ for node in output:
+ if node["node"].type == mm_allocnode_type.pointer():
+ mm_dumpnode(
+ node["node"],
+ node["count"],
+ self.align,
+ simple,
+ detail,
+ self.check_alive(node["pid"]),
+ )
+ else:
+ mempool_dumpbuf(
+ node["node"],
+ node["size"],
+ node["count"],
+ self.align,
+ simple,
+ detail,
+ self.check_alive(node["pid"]),
+ )
+
+ gdb.write("%12s%12s\n" % ("Total Blks", "Total Size"))
+ gdb.write("%12d%12d\n" % (self.aordblks, self.uordblks))
+
def memdump(self, pid, seqmin, seqmax, address, simple, detail,
biggest_top=30):
"""Dump the heap memory"""
@@ -472,6 +500,7 @@ class Memdump(gdb.Command):
title_dict = {
PID_MM_ALLOC: "Dump all used memory node info, use
'\x1b[33;1m*\x1b[m' mark pid does not exist:\n",
+ PID_MM_MEMPOOL: "Dump mempool:\n",
PID_MM_FREE: "Dump all free memory node info:\n",
PID_MM_BIGGEST: f"Dump biggest allocated top {biggest_top}\n",
PID_MM_ORPHAN: "Dump allocated orphan nodes\n",
@@ -505,45 +534,21 @@ class Memdump(gdb.Command):
self.memnode_dump(node)
elif pid == PID_MM_ORPHAN:
for node in alloc_node:
- if node.is_orphan:
+ if node.is_orphan():
+ self.memnode_dump(node)
+ elif pid == PID_MM_MEMPOOL or pid >= 0:
+ for node in alloc_node:
+ if node.pid == pid:
self.memnode_dump(node)
- if not detail:
- output = []
- for node in self.backtrace_dict.values():
- output.append(node)
-
- output.sort(key=get_count, reverse=True)
- for node in output:
- if node["node"].type == mm_allocnode_type.pointer():
- mm_dumpnode(
- node["node"],
- node["count"],
- self.align,
- simple,
- detail,
- self.check_alive(node["pid"]),
- )
- else:
- mempool_dumpbuf(
- node["node"],
- node["size"],
- node["count"],
- self.align,
- simple,
- detail,
- self.check_alive(node["pid"]),
- )
-
- gdb.write("%12s%12s\n" % ("Total Blks", "Total Size"))
- gdb.write("%12d%12d\n" % (self.aordblks, self.uordblks))
+ self.memdump_tail(detail, simple)
def complete(self, text, word):
return gdb.COMPLETE_SYMBOL
def parse_arguments(self, argv):
parser = argparse.ArgumentParser(description="memdump command")
- parser.add_argument("-p", "--pid", type=str, help="Thread PID")
+ parser.add_argument("-p", "--pid", type=str, help="Thread PID, -1 for
mempool")
parser.add_argument("-a", "--addr", type=str, help="Query memory
address")
parser.add_argument("-i", "--min", type=str, help="Minimum value")
parser.add_argument("-x", "--max", type=str, help="Maximum value")
