packages/net/haproxy/patches/0010-BUG-MEDIUM-cli-make-show-fd-thread-safe.patch
Christian Lachner 2cc76214e1 haproxy: Update all patches for HAProxy v1.8.13
- Add new patches (see https://www.haproxy.org/bugs/bugs-1.8.13.html)
- Raise PKG_RELEASE to 2

Signed-off-by: Christian Lachner <gladiac@gmail.com>
2018-08-25 18:51:06 +02:00

50 lines
1.7 KiB
Diff

commit f41ca2546e3c35cc389f45428341ec03dade314d
Author: Willy Tarreau <w@1wt.eu>
Date: Thu Aug 2 11:05:48 2018 +0200
BUG/MEDIUM: cli: make "show fd" thread-safe
The "show fd" command was implemented as a debugging aid but it's not
thread safe. Its features have grown, it can now dump some mux-specific
parts and is being used in production to capture some useful debugging
traces. But it will quickly crash the process when used during an H2 load
test for example, especially when haproxy is built with the DEBUG_UAF
option. It cannot afford not to be thread safe anymore. Let's make use
of the new rendez-vous point using thread_isolate() / thread_release()
to ensure that the data being dumped are not changing under us. The dump
becomes slightly slower under load but now it's safe.
This should be backported to 1.8 along with the rendez-vous point code
once considered stable enough.
(cherry picked from commit bf9fd650883b23604b7cd4aabf04fc0c4c8fe7c7)
Signed-off-by: Willy Tarreau <w@1wt.eu>
diff --git a/src/cli.c b/src/cli.c
index 233c2323..8344fe10 100644
--- a/src/cli.c
+++ b/src/cli.c
@@ -787,10 +787,14 @@ static int cli_io_handler_show_fd(struct appctx *appctx)
void *ctx = NULL;
uint32_t conn_flags = 0;
+ thread_isolate();
+
fdt = fdtab[fd];
- if (!fdt.owner)
+ if (!fdt.owner) {
+ thread_release();
goto skip; // closed
+ }
if (fdt.iocb == conn_fd_handler) {
conn_flags = ((struct connection *)fdt.owner)->flags;
@@ -855,6 +859,8 @@ static int cli_io_handler_show_fd(struct appctx *appctx)
li->bind_conf->frontend->id);
}
+ thread_release();
+
chunk_appendf(&trash, "\n");
if (ci_putchk(si_ic(si), &trash) == -1) {