DM: mevent_add/del refine for Linux

Unlike kqueue/kevent of BSD, the epoll of Linux could be
add/del by using different API on the fly.

This patch drops the notify used by mevent_add/del and
call epoll_ctl to add/delete target fd. Only keeps
global_head to track mevent added and makes the code
logic in mevent_dispatch() a littel bit simpler.

Another thing is related with epoll_ctl. If the target
fd is regular fd which doesn't support epoll, epoll_ctl
will return -1. When DM is start by systemd, the STDIO
is not mapped to terminal, epoll_ctl on STDIO could
return -1. Which block UOS boot. We only call mevent_add
after confirm STDIO is mapped to terminal.

Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Zhao Yakui <yakui.zhao@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
This commit is contained in:
Yin Fengwei 2018-04-10 17:26:39 +08:00 committed by lijinxia
parent 7ed912a088
commit ecf619cb95
3 changed files with 54 additions and 137 deletions

View File

@ -75,13 +75,7 @@ struct mevent {
LIST_ENTRY(mevent) me_list;
};
struct ctl_event {
int op;
int fd;
struct epoll_event ee;
};
static LIST_HEAD(listhead, mevent) global_head, change_head;
static LIST_HEAD(listhead, mevent) global_head;
static void
mevent_qlock(void)
@ -141,80 +135,25 @@ mevent_kq_filter(struct mevent *mevp)
return retval;
}
static int
mevent_kq_flags(struct mevent *mevp)
{
int ret;
switch (mevp->me_state) {
case MEV_ADD:
ret = EPOLL_CTL_ADD; /* implicitly enabled */
break;
case MEV_DEL_PENDING:
ret = EPOLL_CTL_DEL;
break;
default:
assert(0);
break;
}
return ret;
}
static int
mevent_build(int mfd, struct ctl_event *kev)
{
struct mevent *mevp, *tmpp;
int i;
i = 0;
mevent_qlock();
list_foreach_safe(mevp, &change_head, me_list, tmpp) {
if (mevp->me_closefd) {
/*
* A close of the file descriptor will remove the
* event
*/
close(mevp->me_fd);
} else {
kev[i].fd = mevp->me_fd;
kev[i].ee.events = mevent_kq_filter(mevp);
kev[i].op = mevent_kq_flags(mevp);
kev[i].ee.data.ptr = mevp;
i++;
}
mevp->me_cq = 0;
LIST_REMOVE(mevp, me_list);
if (mevp->me_state == MEV_DEL_PENDING)
free(mevp);
else
LIST_INSERT_HEAD(&global_head, mevp, me_list);
assert(i < MEVENT_MAX);
}
mevent_qunlock();
return i;
}
static void
mevent_destroy()
{
struct mevent *mevp, *tmpp;
struct epoll_event ee;
mevent_qlock();
list_foreach_safe(mevp, &global_head, me_list, tmpp) {
LIST_REMOVE(mevp, me_list);
ee.events = mevent_kq_filter(mevp);
ee.data.ptr = mevp;
epoll_ctl(epoll_fd, EPOLL_CTL_DEL, mevp->me_fd, &ee);
if ((mevp->me_type == EVF_READ ||
mevp->me_type == EVF_WRITE)
&& mevp->me_fd != STDIN_FILENO)
close(mevp->me_fd);
LIST_REMOVE(mevp, me_list);
free(mevp);
}
@ -239,6 +178,8 @@ struct mevent *
mevent_add(int tfd, enum ev_type type,
void (*func)(int, enum ev_type, void *), void *param)
{
int ret;
struct epoll_event ee;
struct mevent *lp, *mevp;
if (tfd < 0 || func == NULL)
@ -247,44 +188,41 @@ mevent_add(int tfd, enum ev_type type,
if (type == EVF_TIMER)
return NULL;
mevp = NULL;
mevent_qlock();
/*
* Verify that the fd/type tuple is not present in any list
*/
/* Verify that the fd/type tuple is not present in the list */
LIST_FOREACH(lp, &global_head, me_list) {
if (lp->me_fd == tfd && lp->me_type == type)
goto exit;
}
LIST_FOREACH(lp, &change_head, me_list) {
if (lp->me_fd == tfd && lp->me_type == type)
goto exit;
if (lp->me_fd == tfd && lp->me_type == type) {
mevent_qunlock();
return lp;
}
}
mevent_qunlock();
/*
* Allocate an entry, populate it, and add it to the change list.
* Allocate an entry, populate it, and add it to the list.
*/
mevp = calloc(1, sizeof(struct mevent));
if (mevp == NULL)
goto exit;
return NULL;
mevp->me_fd = tfd;
mevp->me_type = type;
mevp->me_func = func;
mevp->me_param = param;
LIST_INSERT_HEAD(&change_head, mevp, me_list);
mevp->me_cq = 1;
mevp->me_state = MEV_ADD;
mevent_notify();
ee.events = mevent_kq_filter(mevp);
ee.data.ptr = mevp;
ret = epoll_ctl(epoll_fd, EPOLL_CTL_ADD, mevp->me_fd, &ee);
if (ret == 0) {
mevent_qlock();
LIST_INSERT_HEAD(&global_head, mevp, me_list);
mevent_qunlock();
exit:
mevent_qunlock();
return mevp;
return mevp;
} else {
free(mevp);
return NULL;
}
}
int
@ -302,25 +240,20 @@ mevent_disable(struct mevent *evp)
static int
mevent_delete_event(struct mevent *evp, int closefd)
{
struct epoll_event ee;
mevent_qlock();
/*
* Place the entry onto the changed list if not already there, and
* mark as to be deleted.
*/
if (evp->me_cq == 0) {
evp->me_cq = 1;
LIST_REMOVE(evp, me_list);
LIST_INSERT_HEAD(&change_head, evp, me_list);
mevent_notify();
}
evp->me_state = MEV_DEL_PENDING;
if (closefd)
evp->me_closefd = 1;
LIST_REMOVE(evp, me_list);
mevent_qunlock();
ee.events = mevent_kq_filter(evp);
ee.data.ptr = evp;
epoll_ctl(epoll_fd, EPOLL_CTL_DEL, evp->me_fd, &ee);
if (closefd)
close(evp->me_fd);
free(evp);
return 0;
}
@ -364,11 +297,9 @@ mevent_deinit(void)
void
mevent_dispatch(void)
{
struct ctl_event clist[MEVENT_MAX];
struct epoll_event eventlist[MEVENT_MAX];
struct mevent *pipev;
int numev;
int ret;
mevent_tid = pthread_self();
@ -392,24 +323,6 @@ mevent_dispatch(void)
assert(pipev != NULL);
for (;;) {
/*
* Build changelist if required.
* XXX the changelist can be put into the blocking call
* to eliminate the extra syscall. Currently better for
* debug.
*/
int i;
struct epoll_event *e;
numev = mevent_build(epoll_fd, clist);
for (i = 0; i < numev; i++) {
e = &clist[i].ee;
ret = epoll_ctl(epoll_fd, clist[i].op, clist[i].fd, e);
if (ret == -1)
perror("Error return from epoll_ctl");
}
/*
* Block awaiting events
*/
@ -425,5 +338,4 @@ mevent_dispatch(void)
if (vm_get_suspend_mode() != VM_SUSPEND_NONE)
break;
}
mevent_build(epoll_fd, clist);
}

View File

@ -691,12 +691,14 @@ virtio_console_add_backend(struct virtio_console *console,
}
if (virtio_console_backend_can_read(be_type)) {
be->evp = mevent_add(fd, EVF_READ,
virtio_console_backend_read, be);
if (be->evp == NULL) {
WPRINTF(("vtcon: mevent_add failed\n"));
error = -1;
goto out;
if (isatty(fd)) {
be->evp = mevent_add(fd, EVF_READ,
virtio_console_backend_read, be);
if (be->evp == NULL) {
WPRINTF(("vtcon: mevent_add failed\n"));
error = -1;
goto out;
}
}
}

View File

@ -267,8 +267,11 @@ static void
uart_opentty(struct uart_vdev *uart)
{
ttyopen(&uart->tty);
uart->mev = mevent_add(uart->tty.fd, EVF_READ, uart_drain, uart);
assert(uart->mev != NULL);
if (isatty(uart->tty.fd)) {
uart->mev = mevent_add(uart->tty.fd, EVF_READ,
uart_drain, uart);
assert(uart->mev != NULL);
}
}
static uint8_t