Two distinct GFW DNS injector types respond to blocked domains simultaneously, returning duplicate responses per query. They are distinguishable by DNS response flags: the cyclic-ordering injector sets flags 0x8400 (AA bit) when query RD=0, while the second injector sets 0x8180. They also differ in DNS parser behavior: the 0x8180 injector responds to queries using an embedded literal dot (0x2e label separator), while the cyclic injector does not—reversing a behavior documented in the Wallbleed paper.

One type of GFW DNS injector (identified by response flags 0x8400 when query RD=0) uses ordered, cyclic IP address pools. By probing at approximately 100–500 queries per second with a fixed UDP 4-tuple, the complete ordered pool can be recovered: for torproject.org this yields 592 IPv4 addresses and 30 IPv6 addresses. The same pool is documented in Appendix A of the Wallbleed (NDSS 2025) paper for the domain 4.tt.

§Collect DNS responses / §Infer the IP address ordering deployment

Analysis of a June 2025 dataset using DBSCAN clustering confirmed a heterogeneous injector deployment: individual source ports are routed to processes that are either ~100% cyclic or ~100% random-order, with an estimated ratio of approximately 6 cyclic processes to 1 non-linear process. Despite the difference in IP-selection logic, the cyclic and non-linear injectors show virtually identical DNS answer TTL distributions, indicating shared codebase with diverged IP-selection modules rather than entirely separate systems.

mcrmck comment 2025-06-21 evaluation

The GFW DNS injector runs as multiple parallel processes that share the same IP address pool but each maintain an independent incrementing counter. Load balancing to these processes is 4-tuple hash-based (src IP, dst IP, src port, dst port), so a fixed source port is routed deterministically to one process. Probing with variable source ports revealed 3 distinct linear sequences in March 2025, growing to 5 or more by June 2025.

§Visualization (variable source ports) / wkrp comment 2025-06-03 deployment

Between March 26 and June 3, 2025, the GFW cyclic injector's IPv4 pool for torproject.org shrank from 592 to 591 addresses (the IP 162.125.82.7 was removed), and the observable number of load-balanced process sequences increased from 3 to 5. These changes occurred without altering the fundamental cyclic structure or pool membership of the remaining addresses, demonstrating that the injector infrastructure undergoes incremental updates while preserving its core mechanism.

wkrp comment 2025-06-03 (pool diff and process count) evaluation

Wallbleed was a buffer over-read in the GFW's DNS injection subsystem that caused middleboxes to append up to 125 bytes of their own process memory to forged DNS responses. The bug persisted for at least two years (confirmed from October 2021); the GFW issued an incorrect partial patch in November 2023 (Wallbleed v2 remained exploitable) and fully patched it in March 2024. Over 5.1 billion Wallbleed responses were collected during continuous measurement, and an IPv4-wide scan found 242 million IP addresses across 381 autonomous systems receiving Wallbleed-injected responses — including some traffic whose source and destination were both outside China, due to routing through China's network border.

2025-fan-wallbleed §1–§3, §7 dns-poisoningpacket-injection

Iran's DNS censor now injects two distinct block-page IPs: 10.10.34.36 (≈87% of 47,633 censored domains) and 10.10.34.34 (≈13%). Both originate from the same network node at Iran's border. Prior research (Aryan et al. 2013) described only 10.10.34.34. The IP injected correlates strongly with the HTTP censorship method applied: domains with 10.10.34.34 in DNS receive TCP RST via HTTP (86.8% of RST cases), while domains with 10.10.34.36 in DNS receive HTTP block pages (84.6% of block-page cases).

2025-lange-i-ra-nconsistencies §3.1, Table 2 dns-poisoningrst-injectionpacket-injection

Over 2.5 months (Nov 2024–Jan 15, 2025), IRBlock scanned all 11M Iranian IPv4 addresses daily, finding 6.8M IPs subject to DNS poisoning and HTTP blockpage injection, and 5.4M IPs subject to UDP-based traffic disruption. Testing over 700M FQDNs (500M apex domains) revealed 6M banned FQDNs from 3.3M censored apex domains. Of 537 active ASes in Iran, 485 (90.3%) exhibited blocking for at least 25% of assigned IPs. DNS and HTTP censorship overlapped at >99% of censored IPs; UDP blocking was a strict subset of DNS-censored IPs, affecting ~5M addresses.

2025-tai-irblock §5.1, §1 dns-poisoningpacket-injectionhttp3-quic-block

The GFI operates three distinct DNS/HTTP injectors with different fake IP addresses (10.10.34.34, 10.10.34.35, 10.10.34.36) and partially overlapping blocklists—mirroring the GFW's triplet-censor architecture. Injector 10.10.34.35 exhibits TTL reflection (injected response TTL = probe TTL − hop count), identical to the GFW. No IP exclusively receives injections from 10.10.34.34 (a smaller, selective component); the two primary injectors 10.10.34.35 and 10.10.34.36 handle the majority of censorship. Different injectors maintain distinct domain blocklists, meaning which domains a user sees as censored depends on routing through their AS.

2025-tai-irblock §5.4, Table 2 dns-poisoningpacket-injectiondpi

The GFW DNS injector vulnerability enabled reflective amplification attacks with a baseline factor of 4.04× (46-byte payload → 186-byte response). Combined with routing loops — approximately 1,000 destination IP addresses in China were found to loop packets across the GFW more than 30 times, with 159 persisting after two days and a maximum of 119 loop iterations per query — the effective amplification factor reached 481.17×, sufficient to generate 100 Gbps of attack traffic from just over 200 Mbps of source traffic.

2024-sakamoto-bleeding §5.2 Reflective Amplification Attack dns-poisoningpacket-injection

The GFW's DNS packet injector (Injector 3, identified by TTL mirroring and zero IP ID) contained an out-of-bounds read vulnerability: due to missing label-length and null-terminator validation, malformed DNS requests caused the injector to copy adjacent stack memory into forged responses. Over three days in October 2023, researchers collected over 1 TB of data containing over 13 billion leaks, ~87.43% with non-duplicate content, including live Internet traffic transiting China's backbone and stack frames of the GFW's packet-handling processes.

2024-sakamoto-bleeding §3 Vulnerability dns-poisoningpacket-injectiondpi