github/esphome
2
0
Fork 0
mirror of https://github.com/esphome/esphome.git synced 2026-06-24 11:07:33 +00:00
Code Issues Packages Projects Releases Wiki Activity

[socket] DNM: diagnostic logging for #15953 OTA hang on ESP8266

Browse Source 
Add WARN-level traces in LWIPRawImpl::recv_fn and
LWIPRawListenImpl::s_queued_recv_fn so we can confirm whether lwIP is
actually delivering received TCP data up to the application on a
failing OTA. The packet capture in #15953 shows the device's TCP layer
not ACKing the OTA firmware burst even though the same connection's
handshake bytes go through normally and a concurrent API connection
keeps working — these traces are intended to bisect 'is the data
reaching the application's recv callback at all'.

Diagnostic only — not for merge.
This commit is contained in:
J. Nick Koston
2026-05-06 07:44:07 -05:00
parent ff0c5f575e
commit 70047d0948
1 changed files with 21 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
esphome/components/socket/lwip_raw_tcp_impl.cpp
View File

@@ -52,6 +52,12 @@ static const char *const TAG = "socket.lwip";
#define LWIP_LOG(msg, ...)
#endif
// DNM diagnostic: trace lwip TCP receive callback invocations and accept-queue
// activity at WARN level so they show up at default log verbosity. Used to
// diagnose esphome/esphome#15953 (ESP8266 OTA hang where firmware data appears
// on the wire but never reaches the application). Remove before merge.
#define LWIP_DIAG(msg, ...) ESP_LOGW(TAG, "lwip-diag %p: " msg, this, ##__VA_ARGS__)
// Clear arg, recv, and err callbacks, then abort a connected PCB.
// Only valid for full tcp_pcb (not tcp_pcb_listen).
// Must be called before destroying the object that tcp_arg points to —
@@ -432,6 +438,15 @@ err_t LWIPRawImpl::recv_fn(struct pbuf *pb, err_t err) {
// LWIP CALLBACK — runs from IRQ context on RP2040 (low-priority user IRQ).
// No heap allocation allowed — malloc is not IRQ-safe (see #14687).
LWIP_LOG("recv(pb=%p err=%d)", pb, err);
// DNM diagnostic for #15953: prove whether lwip is actually delivering
// OTA firmware bursts up to the application layer. err!=0/pb==nullptr
// are connection-close signals and not part of the symptom we're chasing.
if (pb != nullptr && err == 0) {
LWIP_DIAG("recv tot_len=%u rx_buf=%p rx_off=%u closed=%d", pb->tot_len, this->rx_buf_, this->rx_buf_offset_,
this->rx_closed_ ? 1 : 0);
} else {
LWIP_DIAG("recv close pb=%p err=%d", pb, err);
}
if (err != 0) {
// "An error code if there has been an error receiving Only return ERR_ABRT if you have
// called tcp_abort from within the callback function!"
@@ -736,6 +751,12 @@ err_t LWIPRawListenImpl::s_queued_recv_fn(void *arg, struct tcp_pcb *pcb, struct
// causing the API handshake to silently fail (client sends Hello, server never sees it).
(void) pcb;
auto *entry = reinterpret_cast<QueuedPcb *>(arg);
// DNM diagnostic for #15953.
if (pb != nullptr && err == ERR_OK) {
ESP_LOGW(TAG, "lwip-diag queued-recv entry=%p tot_len=%u rx_buf=%p", entry, pb->tot_len, entry->rx_buf);
} else {
ESP_LOGW(TAG, "lwip-diag queued-recv close entry=%p pb=%p err=%d", entry, pb, err);
}
if (pb == nullptr || err != ERR_OK) {
// Remote closed or error
if (pb != nullptr) {