Date: Sun, 17 Jun 2001 22:28:54 +0200 (MEST) From: Sascha Schumann <sascha@schumann.cx> To: Alfred Perlstein <bright@rush.net> Cc: Valentin Nechayev <netch@iv.nn.kiev.ua>, <freebsd-hackers@FreeBSD.ORG> Subject: Re: poll(2)'s arbitrary limit Message-ID: <Pine.LNX.4.33.0106172157290.6072-200000@rossini.schumann.cx> In-Reply-To: <20010617155740.O1832@superconductor.rush.net>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --]
> > I've been using kern.maxproc=kern.maxprocfiles=2*32768 for my
> > tests and that worked successfully.
>
> Ok, so then we don't need to change FreeBSD?
Well, it would be cool, if an administrator could impose a
more restrictive limit on the maximum number of open fds
while still allowing applications to poll large sets of fds.
For example, the appended patch adds kern.maxfilesperpoll
which is used in place of the current check in poll(). This
could be used to fine-tune servers for specific
configurations more easily. At the same time, it would get
rid of the (per the sys_generic.c comment) "bogus" relation
between fd limits and the limits imposed by poll(2).
- Sascha Experience IRCG
http://schumann.cx/ http://schumann.cx/ircg
[-- Attachment #2 --]
--- ./conf/param.c~ Sun Jun 17 22:04:18 2001
+++ ./conf/param.c Sun Jun 17 22:08:25 2001
@@ -69,6 +69,7 @@
int maxprocperuid = NPROC-1; /* maximum # of processes per user */
int maxfiles = MAXFILES; /* system wide open files limit */
int maxfilesperproc = MAXFILES; /* per-process open files limit */
+int maxfilesperpoll = MAXFILES; /* maximum files per poll */
int ncallout = 16 + NPROC + MAXFILES; /* maximum # of timer events */
int mbuf_wait = 32; /* mbuf sleep time in ticks */
--- ./kern/kern_descrip.c~ Sun Jun 17 22:03:29 2001
+++ ./kern/kern_descrip.c Sun Jun 17 22:05:36 2001
@@ -1454,6 +1454,9 @@
SYSCTL_PROC(_kern, KERN_FILE, file, CTLTYPE_OPAQUE|CTLFLAG_RD,
0, 0, sysctl_kern_file, "S,file", "Entire file table");
+SYSCTL_INT(_kern, KERN_MAXFILESPERPOLL, maxfilesperpoll, CTLFLAG_RW,
+ &maxfilesperpoll, 0, "Maximum files per poll");
+
SYSCTL_INT(_kern, KERN_MAXFILESPERPROC, maxfilesperproc, CTLFLAG_RW,
&maxfilesperproc, 0, "Maximum files allowed open per process");
--- ./kern/sys_generic.c~ Sun Jun 17 22:06:12 2001
+++ ./kern/sys_generic.c Sun Jun 17 22:06:32 2001
@@ -851,14 +851,7 @@
size_t ni;
nfds = SCARG(uap, nfds);
- /*
- * This is kinda bogus. We have fd limits, but that is not
- * really related to the size of the pollfd array. Make sure
- * we let the process use at least FD_SETSIZE entries and at
- * least enough for the current limits. We want to be reasonably
- * safe, but not overly restrictive.
- */
- if (nfds > p->p_rlimit[RLIMIT_NOFILE].rlim_cur && nfds > FD_SETSIZE)
+ if (nfds > maxfilesperpoll)
return (EINVAL);
ni = nfds * sizeof(struct pollfd);
if (ni > sizeof(smallbits))
--- ./sys/file.h~ Sun Jun 17 22:09:53 2001
+++ ./sys/file.h Sun Jun 17 22:10:18 2001
@@ -106,6 +106,7 @@
extern struct fileops badfileops;
extern int maxfiles; /* kernel limit on number of open files */
extern int maxfilesperproc; /* per process limit on number of open files */
+extern int maxfilesperpoll; /* maximum number of fds per poll */
extern int nfiles; /* actual number of open files */
static __inline void fhold __P((struct file *fp));
--- ./sys/sysctl.h~ Sun Jun 17 22:02:18 2001
+++ ./sys/sysctl.h Sun Jun 17 22:10:37 2001
@@ -329,6 +329,7 @@
#define KERN_USRSTACK 33 /* int: address of USRSTACK */
#define KERN_LOGSIGEXIT 34 /* int: do we log sigexit procs? */
#define KERN_MAXID 35 /* number of valid kern ids */
+#define KERN_MAXFILESPERPOLL 36 /* int: max files per poll */
#define CTL_KERN_NAMES { \
{ 0, 0 }, \
@@ -359,6 +360,7 @@
{ "ntp_pll", CTLTYPE_NODE }, \
{ "bootfile", CTLTYPE_STRING }, \
{ "maxfilesperproc", CTLTYPE_INT }, \
+ { "maxfilesperpoll", CTLTYPE_INT }, \
{ "maxprocperuid", CTLTYPE_INT }, \
{ "dumpdev", CTLTYPE_STRUCT }, /* we lie; don't print as int */ \
{ "ipc", CTLTYPE_NODE }, \
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.33.0106172157290.6072-200000>