Date: Mon, 08 May 2017 19:27:24 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-geom@FreeBSD.org Subject: [Bug 198500] [gjournal] sysctl kern.geom.journal.cache.limit 32 bit integer overflow Message-ID: <bug-198500-14739-v2flU7MvOt@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-198500-14739@https.bugs.freebsd.org/bugzilla/> References: <bug-198500-14739@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=198500 longwitz@incore.de changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |longwitz@incore.de --- Comment #1 from longwitz@incore.de --- The main problem caused by the mixed use of u_int and u_long variables in g_journal.c is the suspicious value of kern.geom.journal.cache.limit for machines with more than 8 GB of memory. Therefore I propose the following patch for this problem for FreeBSD 10.3-STABLE #0 r317782: --- sys/geom/journal/g_journal.c.orig 2017-05-04 10:02:46.000000000 +0200 +++ sys/geom/journal/g_journal.c 2017-05-05 23:37:57.625783000 +0200 @@ -131,28 +131,28 @@ SYSCTL_UINT(_kern_geom_journal, OID_AUTO, optimize, CTLFLAG_RW, &g_journal_do_optimize, 0, "Try to combine bios on flush and copy"); -static u_int g_journal_cache_used = 0; -static u_int g_journal_cache_limit = 64 * 1024 * 1024; -TUNABLE_INT("kern.geom.journal.cache.limit", &g_journal_cache_limit); +static u_long g_journal_cache_used = 0; +static u_long g_journal_cache_limit = 64 * 1024 * 1024; +TUNABLE_ULONG("kern.geom.journal.cache.limit", &g_journal_cache_limit); static u_int g_journal_cache_divisor = 2; TUNABLE_INT("kern.geom.journal.cache.divisor", &g_journal_cache_divisor); static u_int g_journal_cache_switch = 90; static u_int g_journal_cache_misses = 0; static u_int g_journal_cache_alloc_failures = 0; -static u_int g_journal_cache_low = 0; +static u_long g_journal_cache_low = 0; static SYSCTL_NODE(_kern_geom_journal, OID_AUTO, cache, CTLFLAG_RW, 0, "GEOM_JOURNAL cache"); -SYSCTL_UINT(_kern_geom_journal_cache, OID_AUTO, used, CTLFLAG_RD, +SYSCTL_ULONG(_kern_geom_journal_cache, OID_AUTO, used, CTLFLAG_RD, &g_journal_cache_used, 0, "Number of allocated bytes"); static int g_journal_cache_limit_sysctl(SYSCTL_HANDLER_ARGS) { - u_int limit; + u_long limit; int error; limit = g_journal_cache_limit; - error = sysctl_handle_int(oidp, &limit, 0, req); + error = sysctl_handle_long(oidp, &limit, 0, req); if (error != 0 || req->newptr == NULL) return (error); g_journal_cache_limit = limit; @@ -160,7 +160,7 @@ return (0); } SYSCTL_PROC(_kern_geom_journal_cache, OID_AUTO, limit, - CTLTYPE_UINT | CTLFLAG_RW, NULL, 0, g_journal_cache_limit_sysctl, "I", + CTLTYPE_ULONG | CTLFLAG_RW, NULL, 0, g_journal_cache_limit_sysctl, "LU", "Maximum number of allocated bytes"); SYSCTL_UINT(_kern_geom_journal_cache, OID_AUTO, divisor, CTLFLAG_RDTUN, &g_journal_cache_divisor, 0, @@ -3035,8 +3035,8 @@ kproc_exit(0); } if (error == 0 && g_journal_sync_requested == 0) { - GJ_DEBUG(1, "Out of cache, force switch (used=%u " - "limit=%u).", g_journal_cache_used, + GJ_DEBUG(1, "Out of cache, force switch (used=%jd " + "limit=%jd).", g_journal_cache_used, g_journal_cache_limit); } GJ_TIMER_START(1, &bt); -- You are receiving this mail because: You are the assignee for the bug.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-198500-14739-v2flU7MvOt>