samples: siwx917_ota: http/s OTAF application #111

Devika-Raju-Z · 2025-09-16T04:54:29Z

Updating the SiWx917 device's firmware including NWP, M4, and the combined M4 + TA images using an HTTP or HTTPS server. It supports both secure and non-secure firmware updates for all three types: NWP, M4, and the combined image.

jerome-pouiller · 2025-09-16T09:14:25Z

samples/siwx917_https_otaf/CMakeLists.txt

+cmake_minimum_required(VERSION 3.20.0)
+
+find_package(Zephyr REQUIRED HINTS $ENV{ZEPHYR_BASE})
+project(siwx917_tcp_client)


OTA already mean Over The Air, so we don't need to specify https. I suggest to rename this sample in siwx91x_ota.

Align the name of the sample with the name of the project.

jerome-pouiller · 2025-09-16T09:22:03Z

samples/siwx917_https_otaf/CMakeLists.txt

+zephyr_code_relocate(FILES ${WISECONNECT_DIR}/components/device/silabs/si91x/wireless/ahb_interface/inc/sli_siwx917_soc.h LOCATION RAM)
+zephyr_code_relocate(FILES ${WISECONNECT_DIR}/components/sli_si91x_wifi_event_handler/src/sli_si91x_wifi_event_handler.c LOCATION RAM)
+zephyr_code_relocate(FILES ${WISECONNECT_DIR}/components/sli_wifi_command_engine/src/sli_wifi_command_engine.c LOCATION RAM)
+zephyr_code_relocate(FILES ${SERVICE_DIR}/sleeptimer/src/sl_sleeptimer.c LOCATION RAM)


Wow, this is very unexpected. Can you explain why it's required? I assume the iea is to relocate in memory all the code reached by the ISR? In this case, these relocations are not sufficient.

Keeping this frequently used files in RAM to improve the performance. We can keep this in flash, but the performance will be low

What performance degradation do you get? If this kind of hack is required to get sane performance, we have raise the issue and address it.

When we relocated the code to RAM for TCP transmit throughput analysis, we observed an improvement of approximately 0.5 Mbps in throughput.

I assume we can consider it is still sane.

jerome-pouiller · 2025-09-16T09:22:20Z

samples/siwx917_https_otaf/README.rst

+Overview
+********
+
+Application demonstrates how to perform HTTP/HTTPS OTA firmware updates on the SiWx917 platform using Zephyr RTOS. It connects to a Wi-Fi network, establishes a secure HTTPS connection using a CA certificate, and downloads firmware updates from a remote server. The application showcases secure connectivity and OTA update mechanisms for IoT devices.


Warp you line at 80 columns.

jerome-pouiller · 2025-09-16T09:24:35Z

samples/siwx917_https_otaf/README.rst

+************
+
+* SiWx917 development board with Wi-Fi support.
+* Windows PC to run HTTP server python script.


Which python script? It does not work with standard file sharing?

This application is verified with HTTP server python script

It is not clear. Can this application run with another http server? If no, why? If yes, why do we require a specific server?

jerome-pouiller · 2025-09-16T09:26:19Z

samples/siwx917_https_otaf/README.rst

+    * For non-secure updates (SECURED_IMAGE_UPDATE=0), the possible values to set HTTP_URL are:
+        * M4_RPS_FILE - M4 image file
+        * TA_RPS_FILE - TA image file
+        * COMBINED_RPS_FILE - Combined image file


I think you can replace app_config.h by Kconfig.

jerome-pouiller · 2025-09-16T10:30:05Z

samples/siwx917_https_otaf/src/main.c

+	int ret;
+
+	ret = setup_socket(family, server, port, sock, addr, addr_len);
+	if (ret < 0 || *sock < 0) {


You will simplify your code if you adopt a more Posix way to design setup_socket(). I suggest to return the socket file descriptor (>= 0) on success and -errno on error.

Ditto for http_connect_socket().

BTW, does it make sense to have setup_socket() and http_connect_socket(). Shouldn't we merge them?

jerome-pouiller · 2025-09-16T10:31:42Z

samples/siwx917_https_otaf/src/main.c

+				printf("\r\nWLAN_SCAN_STATE\r\n");
+				struct wifi_scan_params params = { 0 };
+				params.dwell_time_active = 99;
+				if (net_mgmt(NET_REQUEST_WIFI_SCAN, iface, &params, sizeof(params))) {


Avoid large blocks of code inside switch .. case.

jerome-pouiller · 2025-09-16T10:33:49Z

samples/siwx917_https_otaf/src/main.c

+			}break;
+		}
+	}
+}


I have no clue of what you are doing in this function. I believe the way to manage the state machine is wrong. I believe you can get rid of the states and make the nominal case clearer:

void app_try_ota() { ret = wait_for_network(); if (ret) return; http_connect_socket(); ret = wait_for_server(); if (ret) return; while (end < ota_image_size - 1) { ret = handle_http_packet(msg_c.buffer, msg_c.length); if (ret) return; } sl_si91x_soc_nvic_reset(); } int main() { ... while (1) { // app_try_ota() does not return in case of success app_try_ota(); } }

Using a state machine approach can be effective for managing error conditions. If an error occurs in a particular state, we can handle by transitioning back to the appropriate state.

For now this code is unmaintainable. Unless you know what you are doing, I suggest to change this design.

However, of course, if you are able to present this code properly, I am fine to approve it.

jerome-pouiller · 2025-09-16T10:45:39Z

samples/siwx917_https_otaf/src/main.c

+		}	break;
+	}
+	return status;
+}


Same comment here. Rewrite this function.

Using a state machine approach can be effective for managing error conditions. If an error occurs in a particular state, we can handle by transitioning back to the appropriate state.

jerome-pouiller · 2025-09-16T10:53:39Z

samples/siwx917_https_otaf/src/main.c

+#endif
+	uint32_t status = 0;
+	uint8_t otaf_header_size = 0;
+	switch(otaf_state)


Add a blank line between variable declarations and the first statement.

jerome-pouiller · 2025-09-17T13:29:48Z

samples/siwx917_https_otaf/CMakeLists.txt

+zephyr_code_relocate(FILES ${WISECONNECT_DIR}/components/device/silabs/si91x/wireless/ahb_interface/inc/sli_siwx917_soc.h LOCATION RAM)
+zephyr_code_relocate(FILES ${WISECONNECT_DIR}/components/sli_si91x_wifi_event_handler/src/sli_si91x_wifi_event_handler.c LOCATION RAM)
+zephyr_code_relocate(FILES ${WISECONNECT_DIR}/components/sli_wifi_command_engine/src/sli_wifi_command_engine.c LOCATION RAM)
+zephyr_code_relocate(FILES ${SERVICE_DIR}/sleeptimer/src/sl_sleeptimer.c LOCATION RAM)


What performance degradation do you get? If this kind of hack is required to get sane performance, we have raise the issue and address it.

jerome-pouiller · 2025-09-17T13:31:50Z

samples/siwx917_https_otaf/README.rst

+************
+
+* SiWx917 development board with Wi-Fi support.
+* Windows PC to run HTTP server python script.


It is not clear. Can this application run with another http server? If no, why? If yes, why do we require a specific server?

jerome-pouiller · 2025-09-17T13:49:50Z

samples/siwx917_https_otaf/prj.conf

+# Memory and Threading
+CONFIG_MAIN_STACK_SIZE=2048
+CONFIG_INIT_STACKS=y
+CONFIG_HEAP_MEM_POOL_SIZE=16384


Where this allocation happens? If want to increase the space available for malloc(), this is not the correct option (this option in fact make the memory unavailable for malloc()). This mem pool is normally used for kernel object allocation. Unless you have specific reason, you should have to change it.

jerome-pouiller · 2025-09-17T13:53:21Z

samples/siwx917_https_otaf/prj.conf

+
+# WiFi Configuration
+CONFIG_WIFI=y
+CONFIG_WIFI_SILABS_SIWX91X_NET_STACK_NATIVE=y


I believe this the default and you don't need to specify this option.

I believe this option is to differentiate between native or offload stack. The files siwx91x_wifi.c, siwx91x_wifi_ap.c, and siwx91x_wifi_sta.c contain code that conditionally executes based on the CONFIG_WIFI_SILABS_SIWX91X_NET_STACK_NATIVE macro

Indeed, but CONFIG_WIFI_SILABS_SIWX91X_NET_STACK_NATIVE set by default if you enable Wifi. You don't need to specify it.

jerome-pouiller · 2025-09-17T14:11:04Z

samples/siwx917_https_otaf/src/main.c

+				dhcpv6_params.request_addr = true;
+				dhcpv6_params.request_prefix = false;
+				printf("\r\ndhcpv6 start\r\n");
+				net_dhcpv6_start(iface, &dhcpv6_params);


I think Zephyr support SLAAC for configuration. Then, I don't think you need DHCPv6.

jerome-pouiller · 2025-09-17T14:18:53Z

samples/siwx91x_ota/prj.conf

+CONFIG_MBEDTLS_ENABLE_HEAP=y
+CONFIG_MBEDTLS_HEAP_SIZE=60000
+CONFIG_TLS_CREDENTIALS=y
+CONFIG_NET_SOCKETS_SOCKOPT_TLS=y


I may miss things, but the following config should be very close of what you need ():

CONFIG_NETWORKING=y CONFIG_WIFI=y CONFIG_NET_IPV6=y CONFIG_NET_IPV4=y CONFIG_NET_DHCPV4=y CONFIG_NET_TCP=y CONFIG_NET_SOCKETS=y CONFIG_NET_SOCKETS_SOCKOPT_TLS=y CONFIG_HTTP_CLIENT=y CONFIG_MAIN_STACK_SIZE=2048

You may improve compatibility with TLS servers by adding the snippet below, but I think this is not strictly required.

CONFIG_MBEDTLS_CIPHER_ALL_ENABLED=y CONFIG_MBEDTLS_POLY1305=y

Also note I suggest to use the zsock_*() function so you won't depends on CONFIG_POSIX_API.

jerome-pouiller · 2025-09-17T14:45:27Z

samples/siwx917_https_otaf/src/app_config.h

+#else
+#define SSID                    "<AP-SSID>"
+#define PSK                     "<AP-PASSPHRASE>"
+#define HTTP_SERVER_ADDR        "<SERVER-ADDR>"


hmm... indeed, I see your workaround with TLS_HOSTNAME = NULL. Probably we want to keep that. I suggest to add note explaining this does not for production and how to properly check the FQDN.

Anyway, FQDN will be usual case and we have to accept an hostname instead of an an IP (zsock_getaddrinfo() also accept raw IP address anyway).

jerome-pouiller · 2025-09-17T14:56:47Z

samples/siwx917_https_otaf/src/main.c

+#include "sl_si91x_hal_soc_soft_reset.h"
+
+/* Event masks */
+#define L4_EVENT_MASK (NET_EVENT_L4_IPV6_CONNECTED)


You can just use:

net_mgmt_init_event_callback(&l4_cb, l4_event_handler, NET_EVENT_L4_IPV6_CONNECTED);

L4_EVENT_MASK just make the code more difficult to understand.

jerome-pouiller · 2025-09-17T14:58:12Z

samples/siwx917_https_otaf/src/main.c

+int end = (CHUNK_SIZE - 1);
+int otaf_state = START_IMAGE;
+uint16_t http_sts_code = 0;
+osMessageQueueId_t http_data_q;


osMessageQueueId_t is not a Zephyr native structure. It provided by the CMSIS-RTOS interface to simplify portability of existing applications.

jerome-pouiller · 2025-09-17T14:59:16Z

samples/siwx917_https_otaf/src/main.c

+char link_local[NET_IPV6_ADDR_LEN] = {0};
+char ipv6_address[NET_IPV6_ADDR_LEN] = {0};
+static struct net_mgmt_event_callback l4_cb;
+#endif


Avoid global variables.

jhedberg · 2025-09-26T12:38:51Z

Please rebase this on the latest main branch to see if the Compliance failure gets resolved

jhedberg · 2025-09-26T12:46:37Z

Please rebase this on the latest main branch to see if the Compliance failure gets resolved

Nevermind. Rerunning the workflow was sufficient, since it does a rebase itself. The check is passing now.

jerome-pouiller · 2025-09-26T13:10:42Z

samples/siwx91x_ota/src/app_config.h

+	OTA_STATE_SERVER_CONNECT, /**< Connect to OTA server */
+	OTA_STATE_DOWNLOAD,       /**< Download firmware image */
+	OTA_STATE_PROCESS,        /**< Process downloaded firmware */
+};


This symbol doesn't need to be exported. You can move it to main.c (then app_config.h will be useless)

jerome-pouiller · 2025-09-26T13:14:32Z

samples/siwx91x_ota/src/main.c

+char resolved_addr_str[(CONFIG_OTA_IP_PROTOCOL_SELECTION == 0) ? INET_ADDRSTRLEN
+							       : INET6_ADDRSTRLEN];
+static bool ipv6_addr_config_done;
+volatile enum ota_state state = OTA_STATE_INIT;


Avoid global variables. You can probably move them to struct app_ctx (or even better, change them in local variables).

jerome-pouiller · 2025-09-26T13:16:33Z

samples/siwx91x_ota/src/main.c

+	static struct net_mgmt_event_callback l4_cb;
+	struct app_ctx app_ctx = {.wlan_sem = Z_SEM_INITIALIZER(app_ctx.wlan_sem, 0, 1),
+				  .ota_sem = Z_SEM_INITIALIZER(app_ctx.ota_sem, 0, 1),
+				  .dhcp_sem = Z_SEM_INITIALIZER(app_ctx.dhcp_sem, 0, 1)};


Add new lines after opening brace and before closing braces:

struct app_ctx app_ctx = { .wlan_sem = Z_SEM_INITIALIZER(app_ctx.wlan_sem, 0, 1), .ota_sem = Z_SEM_INITIALIZER(app_ctx.ota_sem, 0, 1), .dhcp_sem = Z_SEM_INITIALIZER(app_ctx.dhcp_sem, 0, 1) };

jerome-pouiller · 2025-09-26T13:17:49Z

samples/siwx91x_ota/src/main.c

+
+	iface = net_if_get_first_wifi();
+	net_mgmt_init_event_callback(&app_ctx.wlan_mgmt_cb, wifi_mgmt_event_handler,
+				     (WIFI_SHELL_MGMT_EVENTS_COMMON));


Parentheses around WIFI_SHELL_MGMT_EVENTS_COMMON are probably useless.

samples/siwx91x_ota/Kconfig

jerome-pouiller · 2025-09-26T15:16:29Z

samples/siwx91x_ota/src/main.c

+
+		case OTA_STATE_IP_CONFIG:
+			ret = configure_ip(app_ctx_st);
+			state = (ret == 0) ? OTA_STATE_SERVER_CONNECT : OTA_STATE_IP_CONFIG;


I am not updating the next state in the dhcp callback.

jerome-pouiller · 2025-09-26T15:19:32Z

samples/siwx91x_ota/src/main.c

+	int firmware_status = 0;
+	char *ssid = CONFIG_OTA_WIFI_SSID;
+	char *pwd = CONFIG_OTA_WIFI_PSK;
+	sl_wifi_firmware_version_t version = {0};


sl_wifi_firmware_version_t version = { }; is sufficient.

jerome-pouiller · 2025-09-26T15:22:33Z

samples/siwx91x_ota/src/main.c

+			} else {
+				printf("Failed to get firmware version: 0x%x\n", ret);
+			}
+			state = OTA_STATE_SCAN;


This state is only used one. You can simplify your FSM by placing this code outside of the loop. (BTW, since version is only used for this part, I suggest to create a function for that task).

jerome-pouiller · 2025-09-26T15:26:03Z

samples/siwx91x_ota/src/main.c

+		break;
+	case NET_EVENT_WIFI_DISCONNECT_RESULT:
+		handle_wifi_disconnect_result(cb);
+		break;


If I understand right, the purpose of these callbacks are only to provide info to the user. In this case, why do they need semaphore? Globally, since the call back only change the state, I am not sure any the semaphore are useful (we only have to ensure state is accessed atomically).

Timeout for receiving the response is implemented using semaphore. If we didn't receive the callback within timeout, status setting as error and retrying.

jerome-pouiller · 2025-09-26T15:44:20Z

samples/siwx91x_ota/src/main.c

+				}
+				/* Purge any remaining messages */
+				k_msgq_purge(&http_data_q);
+				return;


Give an opportunity to the FSM to be useful. Catch the error and repair it. (ditto for wifi disconnection)

jerome-pouiller · 2025-09-30T10:32:00Z

samples/siwx91x_ota/src/main.c

+	char *full_url = CONFIG_OTA_UPDATE_URL;
+	size_t data_len;
+
+	memset(&parser, 0, sizeof(parser));


Redundant with http_parser_url_init().

jerome-pouiller · 2025-09-30T10:34:51Z

samples/siwx91x_ota/src/main.c

+	int ret = http_parser_parse_url(full_url, strlen(full_url), 0, &parser);
+	if (ret) {
+		printf("Failed to parse URL: %s\n", full_url);
+		return -EINVAL;


Considering:

this is a fatal error you can't recover

this is sample (pedagogic) code

You can use __ASSERT(). Same comment in case host, path or schema is not specified.

BTW, you need to retrieve the schema to know if TLS is enabled or not.
You also need to set a default value for the port if it is not specified.

jerome-pouiller · 2025-09-30T10:39:05Z

samples/siwx91x_ota/src/main.c

+		return -errno;
+	}
+
+	if (IS_ENABLED(CONFIG_OTA_HTTPS_SUPPORT)) {


CONFIG_OTA_HTTPS_SUPPORT is not required, you can look at the schema of the URL. You may __ASSERT(IS_ENABLED(CONFIG_NET_SOCKETS_SOCKOPT_TLS), "Application was built without support for TLS");.

jerome-pouiller · 2025-09-30T10:40:09Z

samples/siwx91x_ota/src/main.c

+		addr_ptr = &((struct sockaddr_in6 *)res->ai_addr)->sin6_addr;
+	}
+	zsock_inet_ntop(res->ai_family, addr_ptr, addr_str, sizeof(addr_str));
+	printf("Connected to %s:%s\n", addr_str, http_parse_data_st.port);


Just print http_parse_data_st.host. So you won't have to do the conversion above.

jerome-pouiller · 2025-09-30T10:43:19Z

samples/siwx91x_ota/src/main.c

+		struct dns_resolve_context *dnsctx;
+
+		dnsctx = dns_resolve_get_default();
+		ret = dns_resolve_reconfigure(dnsctx, dns_servers, NULL, DNS_SOURCE_MANUAL);


In Ipv6 DNS is provided by RDNS field of the Router Advertisements. You don't have to manage manually.
You have to add the symbol CONFIG_NET_IPV6_RA_RDNSS in prj.conf.

The user may enable CONFIG_DNS_SERVER_IP_ADDRESSES=y if his network does not support RDNS.

jerome-pouiller · 2025-09-30T10:56:20Z

samples/siwx91x_ota/src/main.c

+	void *addr_ptr;
+	struct zsock_addrinfo hints = {
+		.ai_family = (CONFIG_OTA_IP_PROTOCOL_SELECTION == 0) ? AF_INET : AF_INET6,
+		.ai_socktype = SOCK_STREAM};


Hints is not really required. Zephyr will already automatically selects the right IP version.

jerome-pouiller · 2025-09-30T11:01:42Z

samples/siwx91x_ota/src/main.c

+			printf("Failed to set TLS_HOSTNAME option (%d)\n", -errno);
+			zsock_close(*sock_ptr);
+			zsock_freeaddrinfo(res);
+			return -errno;


I suggest to manage the errors of this function using a label and goto.

BTW, "Connection failed is definitely something we want to report. But, the others (Failed to create TLS socket, Failed to set TLS secure option, etc...) should really not happen. Since it is a pedagogic code, I am not sure it make sense to clutter the code with such debug traces.

jerome-pouiller · 2025-09-30T11:19:59Z

samples/siwx91x_ota/src/main.c

+				   .protocol = "HTTP/1.1",
+				   .response = ota_http_response_cb,
+				   .recv_buf = app_ctx_st->http_recv_buf,
+				   .recv_buf_len = sizeof(app_ctx_st->http_recv_buf)};


Prefer this format:

struct http_request req = { .method = HTTP_GET, .url = app_ctx_st->http_parse_data_st.path, .header_fields = headers, .host = app_ctx_st->http_parse_data_st.host, .protocol = "HTTP/1.1", .response = ota_http_response_cb, .recv_buf = app_ctx_st->http_recv_buf, .recv_buf_len = sizeof(app_ctx_st->http_recv_buf) };

jerome-pouiller · 2025-09-30T11:20:28Z

samples/siwx91x_ota/src/main.c

+{
+	int ret = 0;
+	char range_header[64];
+	const char *headers[2] = {range_header, NULL};


Add space after { and before }.

jerome-pouiller · 2025-09-30T11:23:07Z

samples/siwx91x_ota/src/ca-cert_der.c

+	0x31, 0xee, 0x2a, 0x3b, 0x7b, 0xf0, 0x2c, 0x9e, 0x6f, 0xe9, 0xc4, 0x07, 0x81, 0x24, 0xda,
+	0x05, 0x70, 0x4d, 0xdd, 0x09, 0xae, 0x9e, 0x72, 0xb8, 0x21, 0x0e, 0x8c, 0xb2, 0xab, 0xaa,
+	0x4c, 0x49, 0x10, 0xf7, 0x76, 0xf9, 0xb5, 0x0d, 0x6c, 0x20, 0xd3, 0xdf, 0x7a, 0x06, 0x32,
+	0x8d, 0x29, 0x1f, 0x28, 0x1d, 0x8d, 0x26, 0x33};


The original code with the conversion was fine. Why did you change it?

With .pem format, the client is closing the connection after the TCP handshake. I was unable to connect with server. I checked by enabling the CONFIG_MBEDTLS_PEM_CERTIFICATE_FORMAT, but it didn't work

jerome-pouiller · 2025-09-30T15:11:52Z

samples/siwx91x_ota/src/main.c

+	if (final_data && staging_len > 0) {
+		msg_p.length = staging_len;
+		memcpy(msg_p.buffer, staging_buf, staging_len);
+		if (k_msgq_put(&app_ctx_st->http_data_q, &msg_p, K_NO_WAIT)) {


I am not sure you need this callback. This is only required to process the data on the fly. But in your case, you can wait the request terminate. I believe you just have to process req->recv_buf / req->data_len after the call to http_client_req().

BTW, I believe http_data_q is also useless since we store only one buffer at time and we don't do any new request while the previous one is not finished.

BTW, can you place this function close to download_firmware_chunk(), so the code is easier to read.

We can't use the http_client_req API without passing callback. If we pass callback as NULL, it gives invalid argument error while executing

This application demonstrates the support for OTA firmware upgrade. Signed-off-by: Devika Raju <[email protected]>

Devika-Raju-Z requested a review from a team September 16, 2025 04:54

Devika-Raju-Z force-pushed the https_otaf branch from 466273e to 6a581ad Compare September 16, 2025 05:19

jerome-pouiller requested changes Sep 16, 2025

View reviewed changes

jerome-pouiller reviewed Sep 17, 2025

View reviewed changes

Devika-Raju-Z force-pushed the https_otaf branch 24 times, most recently from 569c7b4 to dd7f3de Compare September 25, 2025 17:03

jhedberg changed the title ~~http/s otaf application~~ samples: siwx917_ota: http/s OTAF application Sep 26, 2025

jerome-pouiller reviewed Sep 26, 2025

View reviewed changes

Devika-Raju-Z force-pushed the https_otaf branch 5 times, most recently from 32f6693 to 053214b Compare September 30, 2025 10:57

jerome-pouiller reviewed Sep 30, 2025

View reviewed changes

Devika-Raju-Z force-pushed the https_otaf branch 5 times, most recently from b8b78eb to 9a126e8 Compare October 1, 2025 08:01

samples: siwx917_ota: http/s OTAF application

8cc35e9

This application demonstrates the support for OTA firmware upgrade. Signed-off-by: Devika Raju <[email protected]>

Devika-Raju-Z force-pushed the https_otaf branch 3 times, most recently from 6025ad2 to 8cc35e9 Compare October 1, 2025 09:07

samples: siwx917_ota: http/s OTAF application #111

Are you sure you want to change the base?

samples: siwx917_ota: http/s OTAF application #111

Uh oh!

Conversation

Devika-Raju-Z commented Sep 16, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jerome-pouiller Sep 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jhedberg commented Sep 26, 2025

Uh oh!

jhedberg commented Sep 26, 2025

Uh oh!

Choose a reason for hiding this comment

jerome-pouiller Sep 16, 2025 •

edited

Loading

Devika-Raju-Z Sep 27, 2025 •

edited

Loading