chromium_netlink_page_size_fix.patch

--- ./net/base/address_tracker_linux.cc.old	2023-03-06 23:53:08.102111285 +0900
+++ ./net/base/address_tracker_linux.cc	2023-03-06 23:53:13.112109094 +0900
@@ -334,7 +334,7 @@
   *address_changed = false;
   *link_changed = false;
   *tunnel_changed = false;
-  char buffer[4096];
+  char buffer[65536];
   bool first_loop = true;
   {
     absl::optional<base::ScopedBlockingCall> blocking_call;

The kernel limits it to min(8k, PAGE_SIZE), so you do not need that big of a buffer. The size is defined in the NLMSG_GOODSIZE macro in include/linux/netlink.h. The kernel netlink code will happily utilize a bigger buffer after the first dump call though.

I'm hoping the real fix contributed is what @asahilina proposed elsewhere:

std::vector buffer(base::GetPageSize());

Juts don't use fixed stack sizes for page size constrained buffers at all, and you fix this problem forever.