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
