Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
On 13/04/15 01:51, Matt Turner wrote: On Sun, Apr 12, 2015 at 6:38 PM, Jason Ekstrand ja...@jlekstrand.net wrote: On Apr 12, 2015 3:24 PM, Thomas Helland thomashellan...@gmail.com wrote: Hi, This looks correct as far as I can tell. I have some comments inline, but I don't feel strongly about either of them, so do as you please. Also, maybe this is a candidate for /src/util ? Maybe. If so I'm OK leaving it here until there are other users. 2015-04-11 2:48 GMT+02:00 Jason Ekstrand ja...@jlekstrand.net: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; Maybe alloced or alloced_mem instead? I was a bit puzzled initially about its purpose. Sure FWIW, I use length and size for these two things. There are also the std::vector names - capacity and size, if anyone is keen on them :-) Jason, Can you squash the following hunk before pushing. diff --git a/src/glsl/Makefile.sources b/src/glsl/Makefile.sources index c471eca..d784a81 100644 --- a/src/glsl/Makefile.sources +++ b/src/glsl/Makefile.sources @@ -22,6 +22,7 @@ NIR_FILES = \ nir/glsl_to_nir.h \ nir/nir.c \ nir/nir.h \ + nir/nir_array.h \ nir/nir_builder.h \ nir/nir_constant_expressions.h \ nir/nir_dominance.c \ Thanks Emil ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
On Thu, Apr 23, 2015 at 10:14 AM, Emil Velikov emil.l.veli...@gmail.com wrote: On 13/04/15 01:51, Matt Turner wrote: On Sun, Apr 12, 2015 at 6:38 PM, Jason Ekstrand ja...@jlekstrand.net wrote: On Apr 12, 2015 3:24 PM, Thomas Helland thomashellan...@gmail.com wrote: Hi, This looks correct as far as I can tell. I have some comments inline, but I don't feel strongly about either of them, so do as you please. Also, maybe this is a candidate for /src/util ? Maybe. If so I'm OK leaving it here until there are other users. 2015-04-11 2:48 GMT+02:00 Jason Ekstrand ja...@jlekstrand.net: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; Maybe alloced or alloced_mem instead? I was a bit puzzled initially about its purpose. Sure FWIW, I use length and size for these two things. There are also the std::vector names - capacity and size, if anyone is keen on them :-) Jason, Can you squash the following hunk before pushing. diff --git a/src/glsl/Makefile.sources b/src/glsl/Makefile.sources index c471eca..d784a81 100644 --- a/src/glsl/Makefile.sources +++ b/src/glsl/Makefile.sources @@ -22,6 +22,7 @@ NIR_FILES = \ nir/glsl_to_nir.h \ nir/nir.c \ nir/nir.h \ + nir/nir_array.h \ nir/nir_builder.h \ nir/nir_constant_expressions.h \ nir/nir_dominance.c \ Sorry, it already got pushed. However, if you put that hunk into a patch, consider it R-B me. --Jason Thanks Emil ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
On Wed, Apr 22, 2015 at 3:03 PM, Connor Abbott cwabbo...@gmail.com wrote: I think it would be better to call this nir_vector rather than nir_array, since this is essentially a poor man's version of std::vector. Is this ok with you? I battled over that one for quite a while. In the end, I settled on nir_array because NIR is a compiler IR for vector architectures and I was concerned that a datatype called nir_vector that was not, in fact, a vector in the mathematical sense would be troublesome. --Jason On Fri, Apr 10, 2015 at 8:48 PM, Jason Ekstrand ja...@jlekstrand.net wrote: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; + void *data; +} nir_array; + +static inline void +nir_array_init(nir_array *arr, void *mem_ctx) +{ + arr-mem_ctx = mem_ctx; + arr-size = 0; + arr-alloc = 0; + arr-data = NULL; +} + +static inline void +nir_array_fini(nir_array *arr) +{ + if (arr-mem_ctx) + ralloc_free(arr-data); + else + free(arr-data); +} + +#define NIR_ARRAY_INITIAL_SIZE 64 + +/* Increments the size of the array by the given ammount and returns a + * pointer to the beginning of the newly added space. + */ +static inline void * +nir_array_grow(nir_array *arr, size_t additional) +{ + size_t new_size = arr-size + additional; + if (new_size arr-alloc) { + if (arr-alloc == 0) + arr-alloc = NIR_ARRAY_INITIAL_SIZE; + + while (new_size arr-alloc) + arr-alloc *= 2; + + if (arr-mem_ctx) + arr-data = reralloc_size(arr-mem_ctx, arr-data, arr-alloc); + else + arr-data = realloc(arr-data, arr-alloc); + } + + void *ptr = (void *)((char *)arr-data + arr-size); + arr-size = new_size; + + return ptr; +} + +#define nir_array_add(arr, type, elem) \ + *(type *)nir_array_grow(arr, sizeof(type)) = (elem) + +#define nir_array_foreach(arr, type, elem) \ + for (type *elem = (type *)(arr)-data; \ +elem (type *)((char *)(arr)-data + (arr)-size); elem++) + +#ifdef __cplusplus +} /* extern C */ +#endif -- 2.3.5 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
I think it would be better to call this nir_vector rather than nir_array, since this is essentially a poor man's version of std::vector. Is this ok with you? On Fri, Apr 10, 2015 at 8:48 PM, Jason Ekstrand ja...@jlekstrand.net wrote: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; + void *data; +} nir_array; + +static inline void +nir_array_init(nir_array *arr, void *mem_ctx) +{ + arr-mem_ctx = mem_ctx; + arr-size = 0; + arr-alloc = 0; + arr-data = NULL; +} + +static inline void +nir_array_fini(nir_array *arr) +{ + if (arr-mem_ctx) + ralloc_free(arr-data); + else + free(arr-data); +} + +#define NIR_ARRAY_INITIAL_SIZE 64 + +/* Increments the size of the array by the given ammount and returns a + * pointer to the beginning of the newly added space. + */ +static inline void * +nir_array_grow(nir_array *arr, size_t additional) +{ + size_t new_size = arr-size + additional; + if (new_size arr-alloc) { + if (arr-alloc == 0) + arr-alloc = NIR_ARRAY_INITIAL_SIZE; + + while (new_size arr-alloc) + arr-alloc *= 2; + + if (arr-mem_ctx) + arr-data = reralloc_size(arr-mem_ctx, arr-data, arr-alloc); + else + arr-data = realloc(arr-data, arr-alloc); + } + + void *ptr = (void *)((char *)arr-data + arr-size); + arr-size = new_size; + + return ptr; +} + +#define nir_array_add(arr, type, elem) \ + *(type *)nir_array_grow(arr, sizeof(type)) = (elem) + +#define nir_array_foreach(arr, type, elem) \ + for (type *elem = (type *)(arr)-data; \ +elem (type *)((char *)(arr)-data + (arr)-size); elem++) + +#ifdef __cplusplus +} /* extern C */ +#endif -- 2.3.5 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
On Wed, Apr 22, 2015 at 6:06 PM, Jason Ekstrand ja...@jlekstrand.net wrote: On Wed, Apr 22, 2015 at 3:03 PM, Connor Abbott cwabbo...@gmail.com wrote: I think it would be better to call this nir_vector rather than nir_array, since this is essentially a poor man's version of std::vector. Is this ok with you? I battled over that one for quite a while. In the end, I settled on nir_array because NIR is a compiler IR for vector architectures and I was concerned that a datatype called nir_vector that was not, in fact, a vector in the mathematical sense would be troublesome. --Jason Ugh, you're right... I guess nir_array would lead to less confusion, even if it's not quite the name that everyone else uses. On Fri, Apr 10, 2015 at 8:48 PM, Jason Ekstrand ja...@jlekstrand.net wrote: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; + void *data; +} nir_array; + +static inline void +nir_array_init(nir_array *arr, void *mem_ctx) +{ + arr-mem_ctx = mem_ctx; + arr-size = 0; + arr-alloc = 0; + arr-data = NULL; +} + +static inline void +nir_array_fini(nir_array *arr) +{ + if (arr-mem_ctx) + ralloc_free(arr-data); + else + free(arr-data); +} + +#define NIR_ARRAY_INITIAL_SIZE 64 + +/* Increments the size of the array by the given ammount and returns a + * pointer to the beginning of the newly added space. + */ +static inline void * +nir_array_grow(nir_array *arr, size_t additional) +{ + size_t new_size = arr-size + additional; + if (new_size arr-alloc) { + if (arr-alloc == 0) + arr-alloc = NIR_ARRAY_INITIAL_SIZE; + + while (new_size arr-alloc) + arr-alloc *= 2; + + if (arr-mem_ctx) + arr-data = reralloc_size(arr-mem_ctx, arr-data, arr-alloc); + else + arr-data = realloc(arr-data, arr-alloc); + } + + void *ptr = (void *)((char *)arr-data + arr-size); + arr-size = new_size; + + return ptr; +} + +#define nir_array_add(arr, type, elem) \ + *(type *)nir_array_grow(arr, sizeof(type)) = (elem) + +#define nir_array_foreach(arr, type, elem) \ + for (type *elem = (type *)(arr)-data; \ +elem (type *)((char *)(arr)-data + (arr)-size); elem++) + +#ifdef __cplusplus +} /* extern C */ +#endif -- 2.3.5 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
On Sun, Apr 12, 2015 at 6:38 PM, Jason Ekstrand ja...@jlekstrand.net wrote: On Apr 12, 2015 3:24 PM, Thomas Helland thomashellan...@gmail.com wrote: Hi, This looks correct as far as I can tell. I have some comments inline, but I don't feel strongly about either of them, so do as you please. Also, maybe this is a candidate for /src/util ? Maybe. If so I'm OK leaving it here until there are other users. 2015-04-11 2:48 GMT+02:00 Jason Ekstrand ja...@jlekstrand.net: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; Maybe alloced or alloced_mem instead? I was a bit puzzled initially about its purpose. Sure FWIW, I use length and size for these two things. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 09/12] nir: Add a simple growing array data structure
On Apr 12, 2015 3:24 PM, Thomas Helland thomashellan...@gmail.com wrote: Hi, This looks correct as far as I can tell. I have some comments inline, but I don't feel strongly about either of them, so do as you please. Also, maybe this is a candidate for /src/util ? Maybe. If so I'm OK leaving it here until there are other users. 2015-04-11 2:48 GMT+02:00 Jason Ekstrand ja...@jlekstrand.net: --- src/glsl/nir/nir_array.h | 96 1 file changed, 96 insertions(+) create mode 100644 src/glsl/nir/nir_array.h diff --git a/src/glsl/nir/nir_array.h b/src/glsl/nir/nir_array.h new file mode 100644 index 000..1db4e8c --- /dev/null +++ b/src/glsl/nir/nir_array.h @@ -0,0 +1,96 @@ +/* + * Copyright © 2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the Software), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + *Jason Ekstrand (ja...@jlekstrand.net) + * + */ + +#pragma once + +#ifdef __cplusplus +extern C { +#endif + +typedef struct { + void *mem_ctx; + size_t size; + size_t alloc; Maybe alloced or alloced_mem instead? I was a bit puzzled initially about its purpose. Sure + void *data; +} nir_array; + +static inline void +nir_array_init(nir_array *arr, void *mem_ctx) +{ + arr-mem_ctx = mem_ctx; + arr-size = 0; + arr-alloc = 0; + arr-data = NULL; +} + +static inline void +nir_array_fini(nir_array *arr) +{ + if (arr-mem_ctx) + ralloc_free(arr-data); + else + free(arr-data); +} + +#define NIR_ARRAY_INITIAL_SIZE 64 + +/* Increments the size of the array by the given ammount and returns a + * pointer to the beginning of the newly added space. + */ A nit, but s/ammount/amount. Yup Maybe also mention that we allocate memory in 2^n increments? Meh. I do that because I happen to think it's nice but it's not at all a requirement. +static inline void * +nir_array_grow(nir_array *arr, size_t additional) +{ + size_t new_size = arr-size + additional; + if (new_size arr-alloc) { + if (arr-alloc == 0) + arr-alloc = NIR_ARRAY_INITIAL_SIZE; + + while (new_size arr-alloc) + arr-alloc *= 2; + + if (arr-mem_ctx) + arr-data = reralloc_size(arr-mem_ctx, arr-data, arr-alloc); + else + arr-data = realloc(arr-data, arr-alloc); Probably should handle failure to allocate and return NULL? Doesn't hurt. I can do that. + } + + void *ptr = (void *)((char *)arr-data + arr-size); + arr-size = new_size; + + return ptr; +} + +#define nir_array_add(arr, type, elem) \ + *(type *)nir_array_grow(arr, sizeof(type)) = (elem) + +#define nir_array_foreach(arr, type, elem) \ + for (type *elem = (type *)(arr)-data; \ +elem (type *)((char *)(arr)-data + (arr)-size); elem++) + +#ifdef __cplusplus +} /* extern C */ +#endif -- 2.3.5 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev