From: Wanlong G. <gao...@cn...> - 2012-10-30 11:03:28
|
On 10/30/2012 06:52 PM, Jan Stancek wrote: > This test picked fixed address in setup() which was used throughout > the test. There has been reports that it can fail with EINVAL. > > I could reproduce it by adding single printf in between > detach/attach: > http://article.gmane.org/gmane.linux.ltp/16480 > > This patch picks suitable base_addr before each test. So even if glibc > would mmap extra page with addr == base_addr picked by setup(), > new suitable address will be chosen. > > Signed-off-by: Jan Stancek <jst...@re...> Pushed, thank you. Wanlong Gao |