Re: [2/2] version: Test for VerQueryValueA

[Stefan Dösinger]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Am 2015-02-18 um 22:41 schrieb Mark Jansen:
> - char preparedbuffer[] = {
> -
>    0x04, 0x01, 0x34, 0x00, 0x00, 0x00, 0x56, 0x00, 0x53, 0x00, 0x5f, 0x00, 0x56, 0x00, 0x45, 0x00,
> -
>    0x52, 0x00, 0x53, 0x00, 0x49, 0x00, 0x4f, 0x00, 0x4e, 0x00, 0x5f, 0x00, 0x49, 0x00, 0x4e, 0x00,
> -
>    0x46, 0x00, 0x4f, 0x00, 0x00, 0x00, 0x00, 0x00, 0xbd, 0x04, 0xef, 0xfe, 0x00, 0x00, 0x01, 0x00,
> -
>    0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00,
>
> ...

Is it possible to put the test data into dlls/version/test/version.rc in plain text instead of using an unreadable blob? If not, please add a comment to explain what is in the blob.
> - ret = VerQueryValueA(preparedbuffer, "StringFileInfo", (LPVOID*)&p, &len);
> - ok(ret, "VerQueryValue error %u\n", GetLastError());
> - ok(len == 0, "VerQueryValue returned %u, expected 0\n", len);
> - ok(p != (char *)0xdeadbeef, "not expected 0xdeadbeef\n");

Does p point to a 0 byte now? (i.e., ok(!*p, "Expected a terminated string\n"); or something like this.)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
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=gsJF
-----END PGP SIGNATURE-----

[2/2] version: Test for VerQueryValueA

[Mark Jansen]

On Mon, Feb 23, 2015 at 9:47 AM, Stefan Dösinger
 wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Am 2015-02-18 um 22:41 schrieb Mark Jansen:
> > - char preparedbuffer[] = {
> > -
> >    0x04, 0x01, 0x34, 0x00, 0x00, 0x00, 0x56, 0x00, 0x53, 0x00, 0x5f, 0x00, 0x56, 0x00, 0x45, 0x00,
> > -
> >    0x52, 0x00, 0x53, 0x00, 0x49, 0x00, 0x4f, 0x00, 0x4e, 0x00, 0x5f, 0x00, 0x49, 0x00, 0x4e, 0x00,
> > -
> >    0x46, 0x00, 0x4f, 0x00, 0x00, 0x00, 0x00, 0x00, 0xbd, 0x04, 0xef, 0xfe, 0x00, 0x00, 0x01, 0x00,
> > -
> >    0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00,
> >
> > ...
>
> Is it possible to put the test data into dlls/version/test/version.rc in plain text instead of using an unreadable blob? If not, please add a comment to explain what is in the blob.

No, this is not possible.
This is simply a version info block (with only one entry in it to keep
it small, but that doesnt influence the test), compiled in the reactos
build environment.
This does create a slightly different resource, but the resource is
accepted by windows resource parser functions, and not by wine. (as
this test shows).
> > - ret = VerQueryValueA(preparedbuffer, "StringFileInfo", (LPVOID*)&p, &len);
> > - ok(ret, "VerQueryValue error %u\n", GetLastError());
> > - ok(len == 0, "VerQueryValue returned %u, expected 0\n", len);
> > - ok(p != (char *)0xdeadbeef, "not expected 0xdeadbeef\n");
>
> Does p point to a 0 byte now? (i.e., ok(!*p, "Expected a terminated string\n"); or something like this.)

No, p points to a data byte, but it can (should) not be read, since
the returned length is 0.
Trying to read it is exactly what is going wrong in the wine implementation,
since it is a non-null byte.
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2
>
> iQIcBAEBAgAGBQJU6ukyAAoJEN0/YqbEcdMwivkQAIz6mLYCOP2607ACoDIuX58M
> 5oaF83RA4wjqTAtOQYbIGFqTtM0kT5rANLgh5cMkgNETSttl8ZgJlyECUmz8ArHA
> VQv0kr0VsgdurFmhoxA37Cw9kou4b2hgsyimfr0dvjahJAroZJgIzXSbtUVgMVQ3
> SUoluqPMSuXod/xW78lVwLI5HLm79/cJdizo+PiWLmLrlNiga6k295eKAfd5Tcco
> qHo+bCf56s0FRlaD4eZ+5AWl5nDP0eP7aGpddBtk4a0Y5E/3npD+9KE2L/hIlTdy
> kbSSW+NUfWAIUKzVFgVr9m/ejGjTMCb6ieHfuK/7S9oU6jCJyM/mwg++vsgdUJbI
> ts03c3UXV/JXGuDSEkxqMNLdhVM0Jv9RLCz4GlwdH7LlSiO9UOGjUpMF2steuULm
> FwirfHXCV1FyZBWIZyHfUw9wykoY4613S49Iq3EGZKKPQmzSi0p+Aygm2Mz6+LhM
> ApI/rVpGs7rqTxYZWZEgcaadpCGlwczOrgbQYEEoxdaYMm8GbEWRyDXjsHiC9NF4
> RbWV+n69+wzC7bOIYqHQSkXEWNEuFlLjn4MElHCqBkGlwTsw9HIPofN6wDbORONt
> xKP5s14VduDyJdT21jYLJJUvIxO3R7OSNxoWIoKBOofFlQeT0c6WHvnqSqW9e7ED
> sX3108dzC77WwIVODqFm
> =gsJF
> -----END PGP SIGNATURE-----