The first error is always the most significant. The following errors usually are the consequences of the first or duplicates.
This introduces single-shot mode, which means that we will print only the first error. It's enabled by default, but user still can switch to the old behavior via 'kasan_print_till_death' boot option. https://jira.sw.ru/browse/PSBM-39339 Signed-off-by: Andrey Ryabinin <aryabi...@odin.com> --- mm/kasan/report.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/mm/kasan/report.c b/mm/kasan/report.c index 17af572..515ab38 100644 --- a/mm/kasan/report.c +++ b/mm/kasan/report.c @@ -216,14 +216,28 @@ static void kasan_report_error(struct kasan_access_info *info) spin_unlock_irqrestore(&report_lock, flags); } +static bool print_till_death; +static int __init kasan_setup(char *arg) +{ + print_till_death = true; + return 0; +} +__setup("kasan_print_till_death", kasan_setup); + void kasan_report(unsigned long addr, size_t size, bool is_write, unsigned long ip) { struct kasan_access_info info; + static bool reported = false; if (likely(!kasan_enabled())) return; + if (likely(!print_till_death)) { + if (reported) + return; + reported = true; + } info.access_addr = (void *)addr; info.access_size = size; info.is_write = is_write; -- 2.4.6 _______________________________________________ Devel mailing list Devel@openvz.org https://lists.openvz.org/mailman/listinfo/devel
