December 2022 - Wine-gitlab

[PATCH v5 0/7] MR51: vkd3d-shader/hlsl: Improve constant handling in copy prop, and extend the use of aoffimmi modifiers.
by Francisco Casas (＠fcasas) 19 Dec '22

19 Dec '22

If a hlsl_ir_load loads a variable whose components are stored from different instructions, copy propagation doesn't replace it. But if all these instructions are constants (which currently is the case for value constructors), the load can be replaced with a constant value, which is what the first patch of this series does. For instance, this shader: ``` sampler s; Texture2D t; float4 main() : sv_target { return t.Gather(s, float2(0.6, 0.6), int2(0, 0)); } ``` results in the following IR before applying the patch: ``` float | 6.00000024e-01 float | 6.00000024e-01 uint | 0 | = (<constructor-2>[@4].x @2) uint | 1 | = (<constructor-2>[@6].x @3) float2 | <constructor-2> int | 0 int | 0 uint | 0 | = (<constructor-5>[@11].x @9) uint | 1 | = (<constructor-5>[@13].x @10) int2 | <constructor-5> float4 | gather_red(resource = t, sampler = s, coords = @8, offset = @15) | return | = (<output-sv_target0> @16) ``` and this IR afterwards: ``` float2 | {6.00000024e-01 6.00000024e-01 } int2 | {0 0 } float4 | gather_red(resource = t, sampler = s, coords = @2, offset = @3) | return | = (<output-sv_target0> @4) ``` This is required to write texel_offsets as aoffimmi modifiers in the sm4 backend, since it expects the texel_offset arguments to be hlsl_ir_constant. This series also: * Allows Gather() methods to use aoffimmi modifiers instead of an additional source register (which is the only way allowed for shader model 4.1), when possible. * Adds support to texel_offsets in the Load() method via aoffimmi modifiers (the only allowed method). -- v5: vkd3d-shader/hlsl: Propagate swizzle chains. vkd3d-shader/hlsl: Replace swizzles with constants in copy prop. tests: Test constant propagation through swizzles. https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/51

3 10

[PATCH 0/2] MR58: vkd3d-shader/hlsl: Reinterpret minimum precision types as their regular counterparts.
by Francisco Casas (＠fcasas) 19 Dec '22

19 Dec '22

Reinterpret min16float, min10float, min16int, min12int, and min16uint as their regular counterparts: float, float, int, int, uint, respectively. A proper implementation would require adding minimum precision indicators to all the dxbc-tpf instructions that use these types. Consider the output of fxc 10.1 with the following shader: ```hlsl uniform int i; float4 main() : sv_target { min16float4 a = {0, 1, 2, i}; min16int2 b = {4, i}; min10float3 c = {6.4, 7, i}; min12int d = 9.4; min16uint4x2 e = {14.4, 15, 16, 17, 18, 19, 20, i}; return mul(e, b) + a + c.xyzx + d; } ``` ``` ps_5_0 dcl_globalFlags refactoringAllowed | enableMinimumPrecision dcl_constantbuffer CB0[1], immediateIndexed dcl_output o0.xyzw dcl_temps 3 imad r0.xyz {min16u}, l(15, 17, 19, 0) {def32 as min16u}, cb0[0].xxxx {def32 as min16u}, l(56, 64, 72, 0) {def32 as min16u} utof r0.xyz {min16f}, r0.xyzx {min16u} imad r1.x {min16u}, cb0[0].x {def32 as min16u}, cb0[0].x {def32 as min16u}, l(80) {def32 as min16u} utof r0.w {min16f}, r1.x {min16u} mov r1.xyz {min16f}, l(0,1.000000,2.000000,0) {def32 as min16f} itof r1.w {min16f}, cb0[0].x add r0.xyzw {min16f}, r0.xyzw {min16f}, r1.xyzw {min16f} mov r2.y {min2_8f}, l(7.000000) {def32 as min2_8f} itof r2.z {min2_8f}, cb0[0].x add r0.yz {min16f}, r0.yyzy {min16f}, r2.yyzy {min2_8f as min16f} add r0.xyzw {min16f}, r0.xyzw {min16f}, l(15.400000, 9.000000, 9.000000, 15.400000) {def32 as min16f} mov o0.xyzw, r0.xyzw {min16f as def32} ret ``` However, if the graphics driver doesn't have minimum precision support, it ignores the minimum precision indicators and runs at 32-bit precision, which is equivalent as working with regular types. -- https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/58

3 5

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

I've given a cursory review to the implementation, and supplied some inline feedback. Besides that, I think it would be a good idea to split up the commit more, and try to apply one fixup at a time. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19811

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

Zebediah Figura (@zfigura) commented about dlls/kernelbase/path.c: > if (!url) > return E_FAIL; > > - srcLen = lstrlenW(url) + 1; > + /* > + * First check for known, valid and typo free scheme > + */ > + for (pos=1; pos<ARRAY_SIZE(url_scheme); pos++) > + { > + len = wcslen(url_scheme[pos]); > + if ( (len <= wcslen(url)) && (!_wcsnicmp(url, url_scheme[pos], len)) ) > + { > + /* > + * check if string fits into maxChars > + */ > + if (len+1 >= maxChars) This comment doesn't seem to tell us anything more than the following line does. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19810

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

Zebediah Figura (@zfigura) commented about dlls/kernelbase/path.c: > + * First check for known, valid and typo free scheme > + */ > + for (pos=1; pos<ARRAY_SIZE(url_scheme); pos++) > + { > + len = wcslen(url_scheme[pos]); > + if ( (len <= wcslen(url)) && (!_wcsnicmp(url, url_scheme[pos], len)) ) > + { > + /* > + * check if string fits into maxChars > + */ > + if (len+1 >= maxChars) > + return S_FALSE; > + > + lstrcpynW(save_str, url_scheme[pos], len+1); > + url += len; > + goto scheme_done; Can we use a helper function to fix up the scheme instead of using a goto? This will probably also help make the function easier to follow, by virtue of being less monolithic. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19809

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

Zebediah Figura (@zfigura) commented about dlls/kernelbase/path.c: > */ > HRESULT WINAPI UrlFixupW(PCWSTR url, PWSTR translatedUrl, DWORD maxChars) > { > - DWORD srcLen; > + DWORD srcLen, dstLen, len; > + DWORD colon_pos = 0, pos = 0; > + wchar_t *helper_str = NULL; > + PWSTR save_str = translatedUrl; We use WCHAR, not wchar_t. It'd also be nice to use a consistent style for all of variable names in this function (wrt snake case vs camel case). In new code I think we prefer snake case. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19808

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

Zebediah Figura (@zfigura) commented about dlls/kernelbase/path.c: > - srcLen = lstrlenW(url) + 1; > + /* > + * First check for known, valid and typo free scheme > + */ > + for (pos=1; pos<ARRAY_SIZE(url_scheme); pos++) > + { > + len = wcslen(url_scheme[pos]); > + if ( (len <= wcslen(url)) && (!_wcsnicmp(url, url_scheme[pos], len)) ) > + { > + /* > + * check if string fits into maxChars > + */ > + if (len+1 >= maxChars) > + return S_FALSE; > + > + lstrcpynW(save_str, url_scheme[pos], len+1); We already know it fits, so this can just be memcpy(). -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19807

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

Zebediah Figura (@zfigura) commented about dlls/kernelbase/path.c: > + DWORD colon_pos = 0, pos = 0; > + wchar_t *helper_str = NULL; > + PWSTR save_str = translatedUrl; > > - FIXME("%s, %p, %ld stub\n", wine_dbgstr_w(url), translatedUrl, maxChars); > + const WCHAR *url_scheme[] = { L"", L"ftp:", L"https:", L"gopher:", L"mailto:", L"news:", > + L"nntp:", L"telnet:", L"wais:", L"file:", L"mk:", > + L"http:", L"shell:", L"snews:", L"local:", > + L"javascript:", L"vbscript:", L"about:", L"res:", > + L"ms-shell-rooted:", L"ms-shell-idlist:", L"hcp:"}; > + > + const WCHAR *url_scheme_short[] = { L"", L"ftp", L"https", L"gopher", L"mailto", L"snews", > + L"nntp", L"telnet", L"wais", L"file", L"mk", > + L"http", L"ms-shell-rooted", L"news", L"local", > + L"javascript", L"vbscript", L"about", L"res", > + L"ms-shell-idlist", L"shell", L"hcp"}; These are constant arrays, so should be "static const WCHAR *const". Also, it seems that one array is the same as the other without the final colon; can we avoid defining two arrays? -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19806

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Zebediah Figura (＠zfigura) 18 Dec '22

18 Dec '22

Zebediah Figura (@zfigura) commented about dlls/kernelbase/path.c: > > -HRESULT WINAPI UrlFixupW(const WCHAR *url, WCHAR *translatedUrl, DWORD maxChars) > +/* > + * from Documentation: > + * https://docs.microsoft.com/en-us/windows/desktop/api/shlwapi/nf-shlwapi-url… > + * > + * UrlFixupW attempts to correct a URL whose protocol identifier is incorrect. > + * For example, htttp will be changed to http. > + * > + * LWSTDAPI UrlFixupW( > + * [in] PCWSTR pcszUrl, > + * [out] PWSTR pszTranslatedUrl, > + * DWORD cchMax > + * ); > +*/ > +HRESULT WINAPI UrlFixupW(PCWSTR url, PWSTR translatedUrl, DWORD maxChars) Without trying to get bogged down in style, this change is somewhat counterproductive; we tend to avoid P* typedefs in new code, and we also have been abandoning the use of "documentation" style headers, which rarely convey any useful information. (This also doesn't really belong in a commit which should only affect tests.) -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19805

1 0

Re: [PATCH v3 0/3] MR1825: kernelbase/path.c: UrlFixupW()
by Thomas Csovcsity (＠thc13) 18 Dec '22

18 Dec '22

On Sun Dec 18 23:01:30 2022 +0000, Zebediah Figura wrote: > Hi Thomas, thanks for the patch! > We want the tests to "pass" after every commit, i.e. have no failures. > That may mean, in this case, introducing the tests marked as todo, and > then later removing those todos. Ok. Got it. Will do it soon -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1825#note_19804

1 0