Don't SIGABORT if device-mapper doesn't generate uevent for DeleteDevice
It looks like in some rare scenarios device-mapper won't create an
uevent while handling DM_DEV_REMOVE ioctl.
Since failure to generate uevent will result in ueventd not cleaning up
/dev paths in user space, it should be ok to just fail DeleteDevice call
without aborting the entire process.
Test: presubmit
Bug: 171266986
Change-Id: Iece28e6e02c35ac4bf33b9be74dfe1bbbb466bd4
diff --git a/fs_mgr/libdm/dm.cpp b/fs_mgr/libdm/dm.cpp
index bb7d8b3..d5b8a61 100644
--- a/fs_mgr/libdm/dm.cpp
+++ b/fs_mgr/libdm/dm.cpp
@@ -111,8 +111,10 @@
// Check to make sure appropriate uevent is generated so ueventd will
// do the right thing and remove the corresponding device node and symlinks.
- CHECK(io.flags & DM_UEVENT_GENERATED_FLAG)
- << "Didn't generate uevent for [" << name << "] removal";
+ if ((io.flags & DM_UEVENT_GENERATED_FLAG) == 0) {
+ LOG(ERROR) << "Didn't generate uevent for [" << name << "] removal";
+ return false;
+ }
if (timeout_ms <= std::chrono::milliseconds::zero()) {
return true;