Mercurial > hg > nginx-quic
view src/http/ngx_http_busy_lock.h @ 5594:c8a14fbd3ce9 stable-1.4
Upstream: ngx_post_event() instead of upgraded call (ticket #503).
If a request is finalized in the first call to the
ngx_http_upstream_process_upgraded() function, e.g., because upstream
server closed the connection for some reason, in the second call
the u->peer.connection pointer will be null, resulting in segmentation
fault.
Fix is to avoid second direct call, and post event instead. This ensures
that ngx_http_upstream_process_upgraded() won't be called again if
a request is finalized.
author | Maxim Dounin <mdounin@mdounin.ru> |
---|---|
date | Tue, 18 Feb 2014 17:30:40 +0400 |
parents | d620f497c50f |
children |
line wrap: on
line source
/* * Copyright (C) Igor Sysoev * Copyright (C) Nginx, Inc. */ #ifndef _NGX_HTTP_BUSY_LOCK_H_INCLUDED_ #define _NGX_HTTP_BUSY_LOCK_H_INCLUDED_ #include <ngx_config.h> #include <ngx_core.h> #include <ngx_event.h> #include <ngx_http.h> typedef struct { u_char *md5_mask; char *md5; int cacheable; int busy; int max_busy; int waiting; int max_waiting; time_t timeout; ngx_event_mutex_t *mutex; } ngx_http_busy_lock_t; typedef struct { time_t time; ngx_event_t *event; void (*event_handler)(ngx_event_t *ev); u_char *md5; int slot; } ngx_http_busy_lock_ctx_t; int ngx_http_busy_lock(ngx_http_busy_lock_t *bl, ngx_http_busy_lock_ctx_t *bc); int ngx_http_busy_lock_cacheable(ngx_http_busy_lock_t *bl, ngx_http_busy_lock_ctx_t *bc, int lock); void ngx_http_busy_unlock(ngx_http_busy_lock_t *bl, ngx_http_busy_lock_ctx_t *bc); char *ngx_http_set_busy_lock_slot(ngx_conf_t *cf, ngx_command_t *cmd, void *conf); #endif /* _NGX_HTTP_BUSY_LOCK_H_INCLUDED_ */