[PATCH 1/5] msi/tests: Add some tests for remote MSI summary info APIs.

[Dmitry Timoshkov]

Signed-off-by: Dmitry Timoshkov 
---
 dlls/msi/tests/custom.c | 52 ++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 51 insertions(+), 1 deletion(-)

diff --git a/dlls/msi/tests/custom.c b/dlls/msi/tests/custom.c
index 6332d68768..1cf24536d0 100644
--- a/dlls/msi/tests/custom.c
+++ b/dlls/msi/tests/custom.c
@@ -30,6 +30,7 @@
 #include <msxml.h>
 #include <msi.h>
 #include <msiquery.h>
+#include <msidefs.h>

static int todo_level, todo_do_loop;

@@ -279,7 +280,9 @@ static void test_db(MSIHANDLE hinst)
     MSIHANDLE hdb, view, rec, rec2, suminfo;
     char buffer[10];
     DWORD sz;
-    UINT r;
+    UINT r, count, type;
+    INT int_value;
+    FILETIME ft;

hdb = MsiGetActiveDatabase(hinst);
     ok(hinst, hdb, "MsiGetActiveDatabase failed\n");
@@ -452,6 +455,53 @@ static void test_db(MSIHANDLE hinst)
     r = MsiGetSummaryInformationA(hdb, NULL, 1, &suminfo);
     ok(hinst, !r, "got %u\n", r);

+    r = MsiSummaryInfoGetPropertyCount(suminfo, NULL);
+todo_wine
+    ok(hinst, r == RPC_X_NULL_REF_POINTER, "got %u\n", r);
+
+    count = 0xdeadbeef;
+    r = MsiSummaryInfoGetPropertyCount(suminfo, &count);
+todo_wine
+    ok(hinst, !r, "got %u\n", r);
+todo_wine
+    ok(hinst, count == 5, "got %u\n", count);
+
+    r = MsiSummaryInfoGetPropertyA(suminfo, 0, NULL, NULL, NULL, NULL, NULL);
+todo_wine
+    ok(hinst, r == RPC_X_NULL_REF_POINTER, "got %u\n", r);
+
+    type = 0xdeadbeef;
+    int_value = 0xdeadbeef;
+    strcpy(buffer, "deadbeef");
+    sz = sizeof(buffer);
+    r = MsiSummaryInfoGetPropertyA(suminfo, PID_AUTHOR, &type, &int_value, &ft, buffer, &sz);
+todo_wine
+    ok(hinst, !r, "got %u\n", r);
+todo_wine
+    ok(hinst, type == 0, "got %u\n", type);
+todo_wine
+    ok(hinst, int_value == 0, "got %u\n", int_value);
+    ok(hinst, sz == sizeof(buffer), "got %u\n", sz);
+    ok(hinst, !lstrcmpA(buffer, "deadbeef"), "got %s\n", buffer);
+
+    type = 0xdeadbeef;
+    int_value = 0xdeadbeef;
+    strcpy(buffer, "deadbeef");
+    sz = sizeof(buffer);
+    r = MsiSummaryInfoGetPropertyA(suminfo, PID_CODEPAGE, &type, &int_value, &ft, buffer, &sz);
+todo_wine
+    ok(hinst, !r, "got %u\n", r);
+todo_wine
+    ok(hinst, type == 0, "got %u\n", type);
+todo_wine
+    ok(hinst, int_value == 0, "got %u\n", int_value);
+    ok(hinst, sz == sizeof(buffer), "got %u\n", sz);
+    ok(hinst, !lstrcmpA(buffer, "deadbeef"), "got %s\n", buffer);
+
+    r = MsiSummaryInfoSetPropertyA(suminfo, PID_CODEPAGE, VT_I2, 1252, &ft, "");
+todo_wine
+    ok(hinst, r == ERROR_FUNCTION_FAILED, "got %u\n", r);
+
     r = MsiCloseHandle(suminfo);
     ok(hinst, !r, "got %u\n", r);

[Marvin]

Hi,

While running your changed tests on Windows, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=41341

Your paranoid android.

=== w7u (32 bit action) ===
The test timed out

=== w7u (32 bit install) ===
The test timed out

=== w7u (32 bit msi) ===
The test timed out

[Dmitry Timoshkov]

Hello,

Is there any responsibility of a reviewer or a reasonable time frame
for a review, especially such a trivial patch set?

[Hans Leidekker]

On Thu, 2018-09-06 at 23:13 +0800, Dmitry Timoshkov wrote:
> Is there any responsibility of a reviewer or a reasonable time frame
> for a review, especially such a trivial patch set?

You got a comment from Nikolay. I think he has a point, but I'm not
clear on what it takes to handle SEH properly for these RPC calls.

If it's not too much effort I would prefer doing that instead of
spreading the current hack.

[Zebediah Figura]

On 06/09/18 10:34, Hans Leidekker wrote:
> On Thu, 2018-09-06 at 23:13 +0800, Dmitry Timoshkov wrote:
> > Is there any responsibility of a reviewer or a reasonable time frame
> > for a review, especially such a trivial patch set?
>
> You got a comment from Nikolay. I think he has a point, but I'm not
> clear on what it takes to handle SEH properly for these RPC calls.
>
> If it's not too much effort I would prefer doing that instead of
> spreading the current hack.

I believe it's just a matter of wrapping each function call in a
try/catch block.

Since this is effectively my fault, I'd be willing to handle this.

[Dmitry Timoshkov]

Hans Leidekker  wrote:
> On Thu, 2018-09-06 at 23:13 +0800, Dmitry Timoshkov wrote:
> > Is there any responsibility of a reviewer or a reasonable time frame
> > for a review, especially such a trivial patch set?
>
> You got a comment from Nikolay. I think he has a point, but I'm not
> clear on what it takes to handle SEH properly for these RPC calls.

Nikolay's comment is completely misleading and misguided.
> If it's not too much effort I would prefer doing that instead of
> spreading the current hack.

I just followed current pattern of handling NULL pointers. If there's
a consensus that it needs a different approach then all the APIs should
be reworked at once.