[PATCH] msvcp140/tests: Don't trace the current time.

[Francois Gouget]

This causes the failure to always look new and the last_write_time 
value does not really seem useful.

Signed-off-by: Francois Gouget 
---

This also only fails on my box, but it does not change the fact that a 
failure will always look new.

dlls/msvcp140/tests/msvcp140.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/dlls/msvcp140/tests/msvcp140.c b/dlls/msvcp140/tests/msvcp140.c
index 751b1beed86..35a650ddbae 100644
--- a/dlls/msvcp140/tests/msvcp140.c
+++ b/dlls/msvcp140/tests/msvcp140.c
@@ -1239,8 +1239,7 @@ static void test_Last_write_time(void)
     newtime = last_write_time + 222222;
     p_Set_last_write_time(f1W, newtime);
     ok(last_write_time != p_Last_write_time(f1W),
-            "last_write_time should have changed: %s\n",
-            wine_dbgstr_longlong(last_write_time));
+       "last_write_time should have changed\n");

/* test the formula */
     file = CreateFileW(f1W, 0, FILE_SHARE_DELETE | FILE_SHARE_READ | FILE_SHARE_WRITE,

[Piotr Caban]

On 12/30/19 12:06 AM, Francois Gouget wrote:
> This causes the failure to always look new and the last_write_time
> value does not really seem useful.
> Signed-off-by: Francois Gouget
>
> This also only fails on my box, but it does not change the fact that a
> failure will always look new.
>
> dlls/msvcp140/tests/msvcp140.c | 3 +--
>   1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/dlls/msvcp140/tests/msvcp140.c b/dlls/msvcp140/tests/msvcp140.c
> index 751b1beed86..35a650ddbae 100644
> --- a/dlls/msvcp140/tests/msvcp140.c
> +++ b/dlls/msvcp140/tests/msvcp140.c
> @@ -1239,8 +1239,7 @@ static void test_Last_write_time(void)
>       newtime = last_write_time + 222222;
>       p_Set_last_write_time(f1W, newtime);
>       ok(last_write_time != p_Last_write_time(f1W),
> -
>        "last_write_time should have changed: %s\n",
> -
>        wine_dbgstr_longlong(last_write_time));
> -
>   "last_write_time should have changed\n");

Is the failure caused by filesystem write time resolution? Is it fixed 
by changing 222222 to something bigger than 2 seconds (e.g. 22222200)? 
What filesystem is it not working on?

Thanks,
Piotr

[Francois Gouget]

On Mon, 30 Dec 2019, Piotr Caban wrote:
[...]
> Is the failure caused by filesystem write time resolution? Is it fixed by
> changing 222222 to something bigger than 2 seconds (e.g. 22222200)?

I applied the patch below but it did not make a difference.
msvcp140.c:1241: Test failed: last_write_time should have changed: 380d54888384b2

--- a/dlls/msvcp140/tests/msvcp140.c
+++ b/dlls/msvcp140/tests/msvcp140.c
@@ -1236,7 +1236,7 @@ static void test_Last_write_time(void)
     CloseHandle(file);

last_write_time = p_Last_write_time(f1W);
-    newtime = last_write_time + 222222;
+    newtime = last_write_time + 22222200;
     p_Set_last_write_time(f1W, newtime);
     ok(last_write_time != p_Last_write_time(f1W),
             "last_write_time should have changed: %s\n",
> What filesystem is it not working on?

It's an NTFS filesystem.
Here is more information about the VM:

Description=French Windows XP SP3: Updated on 2016/12/03, added 
Silverlight, optional DirectX 9.0c components, MSXML 4sp3, Visual C++ 
2005sp1, 2008sp1, 2010sp1, 2012u4, 2013, and 2015 runtimes. 
CPU:haswell-notsx*2cores, RAM:1GB, Disk:virtio, ethernet:virtio, 
sound:ac97, GPU:qxl, display:spice, screens: 2, TestAgentd:detached
    dwMajorVersion=5
    dwMinorVersion=1
    dwBuildNumber=2600
    PlatformId=2
    szCSDVersion=Service Pack 3
    wServicePackMajor=3
    wServicePackMinor=0
    wSuiteMask=768
    wProductType=1
    wReserved=0
    SystemDefaultLCID=040c
    UserDefaultLCID=040c
    ThreadLocale=040c
    UserDefaultUILanguage=040c