From mboxrd@z Thu Jan 1 00:00:00 1970 From: gorcunov at gmail.com (Cyrill Gorcunov) Date: Fri, 9 Nov 2018 14:41:23 +0300 Subject: [PATCH] proc: fix and merge proc-self-map-file tests In-Reply-To: <20181109113036.21512-1-rafael.tinoco@linaro.org> References: <20181108111105.GC13195@uranus.lan> <20181109113036.21512-1-rafael.tinoco@linaro.org> Message-ID: <20181109114123.GK13195@uranus.lan> On Fri, Nov 09, 2018 at 09:30:36AM -0200, Rafael David Tinoco wrote: > Merge proc-self-map-files tests into one since this test should focus in > testing readlink in /proc/self/map_files/* only, and not trying to test > mapping virtual address 0. > > Lowest virtual address for user space mapping in other architectures, > like arm, is *at least* *(PAGE_SIZE * 2) and NULL hint does not > guarantee that when MAP_FIXED flag, important to this test, is given. > This patch also fixes this issue in remaining test. > > Link: https://bugs.linaro.org/show_bug.cgi?id=3782 > Signed-off-by: Rafael David Tinoco Wait, Rafael. But we will loose the test of mapping virtual address 0 then. I though you would move testing of virtual address 0 into a separate testcase. I mean testing of first page was a positive side effect of the former Alexey's patch, so we definitely should keep it on x86 at least. Gimme some time I'll try to address it today evening or tomorrow. I think this way everybody will be happy: procfs get passed on arm32 and x86 will still have first page testing. From mboxrd@z Thu Jan 1 00:00:00 1970 From: gorcunov@gmail.com (Cyrill Gorcunov) Date: Fri, 9 Nov 2018 14:41:23 +0300 Subject: [PATCH] proc: fix and merge proc-self-map-file tests In-Reply-To: <20181109113036.21512-1-rafael.tinoco@linaro.org> References: <20181108111105.GC13195@uranus.lan> <20181109113036.21512-1-rafael.tinoco@linaro.org> Message-ID: <20181109114123.GK13195@uranus.lan> Content-Type: text/plain; charset="UTF-8" Message-ID: <20181109114123.73fKrMU8VVeUWCEqtGiGsInkvDQ4_LB1Zmc0f2-d8jM@z> On Fri, Nov 09, 2018@09:30:36AM -0200, Rafael David Tinoco wrote: > Merge proc-self-map-files tests into one since this test should focus in > testing readlink in /proc/self/map_files/* only, and not trying to test > mapping virtual address 0. > > Lowest virtual address for user space mapping in other architectures, > like arm, is *at least* *(PAGE_SIZE * 2) and NULL hint does not > guarantee that when MAP_FIXED flag, important to this test, is given. > This patch also fixes this issue in remaining test. > > Link: https://bugs.linaro.org/show_bug.cgi?id=3782 > Signed-off-by: Rafael David Tinoco Wait, Rafael. But we will loose the test of mapping virtual address 0 then. I though you would move testing of virtual address 0 into a separate testcase. I mean testing of first page was a positive side effect of the former Alexey's patch, so we definitely should keep it on x86 at least. Gimme some time I'll try to address it today evening or tomorrow. I think this way everybody will be happy: procfs get passed on arm32 and x86 will still have first page testing.