15.0.0: add 72_all_PR117177-libcpp.patch

Fixes building glibc. Approved upstream but not yet merged.

Signed-off-by: Sam James <sam@gentoo.org>

1 files changed, 234 insertions, 0 deletions

diff --git a/15.0.0/gentoo/72_all_PR117177-libcpp.patch b/15.0.0/gentoo/72_all_PR117177-libcpp.patch
new file mode 100644
index 0000000..1db42c9
--- /dev/null
+++ b/15.0.0/gentoo/72_all_PR117177-libcpp.patch
@@ -0,0 +1,234 @@
+From b2681b59424dd3a5fdf4101f8cc9eff60a5166b7 Mon Sep 17 00:00:00 2001
+Message-ID: <b2681b59424dd3a5fdf4101f8cc9eff60a5166b7.1729138476.git.sam@gentoo.org>
+From: Jakub Jelinek <jakub@redhat.com>
+Date: Wed, 16 Oct 2024 23:09:32 +0200
+Subject: [PATCH] c: Fix up speed up compilation of large char array
+ initializers when not using #embed [PR117177]
+
+Hi!
+
+Apparently my
+c: Speed up compilation of large char array initializers when not using #embed
+patch broke building glibc.
+
+The issue is that when using CPP_EMBED, we are guaranteed by the
+preprocessor that there is CPP_NUMBER CPP_COMMA before it and
+CPP_COMMA CPP_NUMBER after it (or CPP_COMMA CPP_EMBED), so RAW_DATA_CST
+never ends up at the end of arrays of unknown length.
+Now, the c_parser_initval optimization attempted to preserve that property
+rather than changing everything that e.g. inferes array number of elements
+from the initializer etc. to deal with RAW_DATA_CST at the end, but
+it didn't take into account the possibility that there could be
+CPP_COMMA followed by CPP_CLOSE_BRACE (where the CPP_COMMA is redundant).
+
+As we are peaking already at 4 tokens in that code, peeking more would
+require using raw tokens and that seems to be expensive doing it for
+every pair of tokens due to vec_free done when we are out of raw tokens.
+
+So, the following patch instead determines the case where we want
+another INTEGER_CST element after it after consuming the tokens, and just
+arranges for another process_init_element.
+
+Ok for trunk if this passes bootstrap/regtest?
+
+2024-10-16  Jakub Jelinek  <jakub@redhat.com>
+
+	PR c/117177
+gcc/c/
+	* c-parser.cc (c_parser_initval): Instead of doing
+	orig_len == INT_MAX checks before consuming tokens to set
+	last = 1, check it after consuming it and if not followed
+	by CPP_COMMA CPP_NUMBER, call process_init_element once
+	more with the last CPP_NUMBER.
+gcc/testsuite/
+	* c-c++-common/init-4.c: New test.
+---
+ gcc/c/c-parser.cc                   | 35 ++++++++---
+ gcc/testsuite/c-c++-common/init-4.c | 97 +++++++++++++++++++++++++++++
+ 2 files changed, 122 insertions(+), 10 deletions(-)
+ create mode 100644 gcc/testsuite/c-c++-common/init-4.c
+
+diff --git a/gcc/c/c-parser.cc b/gcc/c/c-parser.cc
+index e4381044e5c..090ab1cbc08 100644
+--- a/gcc/c/c-parser.cc
++++ b/gcc/c/c-parser.cc
+@@ -6529,6 +6529,7 @@ c_parser_initval (c_parser *parser, struct c_expr *after,
+ 	unsigned int i;
+ 	gcc_checking_assert (len >= 64);
+ 	location_t last_loc = UNKNOWN_LOCATION;
++	location_t prev_loc = UNKNOWN_LOCATION;
+ 	for (i = 0; i < 64; ++i)
+ 	  {
+ 	    c_token *tok = c_parser_peek_nth_token_raw (parser, 1 + 2 * i);
+@@ -6544,6 +6545,7 @@ c_parser_initval (c_parser *parser, struct c_expr *after,
+ 	    buf1[i] = (char) tree_to_uhwi (tok->value);
+ 	    if (i == 0)
+ 	      loc = tok->location;
++	    prev_loc = last_loc;
+ 	    last_loc = tok->location;
+ 	  }
+ 	if (i < 64)
+@@ -6567,6 +6569,7 @@ c_parser_initval (c_parser *parser, struct c_expr *after,
+ 	unsigned int max_len = 131072 - offsetof (struct tree_string, str) - 1;
+ 	unsigned int orig_len = len;
+ 	unsigned int off = 0, last = 0;
++	unsigned char lastc = 0;
+ 	if (!wi::neg_p (wi::to_wide (val)) && wi::to_widest (val) <= UCHAR_MAX)
+ 	  off = 1;
+ 	len = MIN (len, max_len - off);
+@@ -6596,20 +6599,25 @@ c_parser_initval (c_parser *parser, struct c_expr *after,
+ 	    if (tok2->type != CPP_COMMA && tok2->type != CPP_CLOSE_BRACE)
+ 	      break;
+ 	    buf2[i + off] = (char) tree_to_uhwi (tok->value);
+-	    /* If orig_len is INT_MAX, this can be flexible array member and
+-	       in that case we need to ensure another element which
+-	       for CPP_EMBED is normally guaranteed after it.  Include
+-	       that byte in the RAW_DATA_OWNER though, so it can be optimized
+-	       later.  */
+-	    if (tok2->type == CPP_CLOSE_BRACE && orig_len == INT_MAX)
+-	      {
+-		last = 1;
+-		break;
+-	      }
++	    prev_loc = last_loc;
+ 	    last_loc = tok->location;
+ 	    c_parser_consume_token (parser);
+ 	    c_parser_consume_token (parser);
+ 	  }
++	/* If orig_len is INT_MAX, this can be flexible array member and
++	   in that case we need to ensure another element which
++	   for CPP_EMBED is normally guaranteed after it.  Include
++	   that byte in the RAW_DATA_OWNER though, so it can be optimized
++	   later.  */
++	if (orig_len == INT_MAX
++	    && (!c_parser_next_token_is (parser, CPP_COMMA)
++		|| c_parser_peek_2nd_token (parser)->type != CPP_NUMBER))
++	  {
++	    --i;
++	    last = 1;
++	    std::swap (prev_loc, last_loc);
++	    lastc = (unsigned char) buf2[i + off];
++	  }
+ 	val = make_node (RAW_DATA_CST);
+ 	TREE_TYPE (val) = integer_type_node;
+ 	RAW_DATA_LENGTH (val) = i;
+@@ -6625,6 +6633,13 @@ c_parser_initval (c_parser *parser, struct c_expr *after,
+ 	init.original_type = integer_type_node;
+ 	init.m_decimal = 0;
+ 	process_init_element (loc, init, false, braced_init_obstack);
++	if (last)
++	  {
++	    init.value = build_int_cst (integer_type_node, lastc);
++	    init.original_code = INTEGER_CST;
++	    set_c_expr_source_range (&init, prev_loc, prev_loc);
++	    process_init_element (prev_loc, init, false, braced_init_obstack);
++	  }
+       }
+ }
+ 
+diff --git a/gcc/testsuite/c-c++-common/init-4.c b/gcc/testsuite/c-c++-common/init-4.c
+new file mode 100644
+index 00000000000..d575a3f5a51
+--- /dev/null
++++ b/gcc/testsuite/c-c++-common/init-4.c
+@@ -0,0 +1,97 @@
++/* { dg-do run } */
++/* { dg-options "-O2" } */
++
++unsigned char a1[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00,
++  0xfa, 0xfb, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
++};
++unsigned char a2[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00
++};
++unsigned char a3[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00,
++};
++unsigned char a4[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00,
++  0xfa
++};
++unsigned char a5[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00,
++  0xfa,
++};
++unsigned char a6[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00,
++  0xfa, 0xfb
++};
++unsigned char a7[] = {
++  0xc8, 0xc9, 0xca, 0xcb, 0x00, 0xcd, 0xce, 0xcf,
++  0x00, 0xd1, 0x00, 0xd3, 0xd4, 0x00, 0xd6, 0xd7,
++  0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0x00, 0x00, 0xdf,
++  0xe0, 0xe1, 0xe2, 0x00, 0xe4, 0xe5, 0xe6, 0xe7,
++  0xe8, 0xe9, 0xea, 0xeb, 0x00, 0xed, 0xee, 0xef,
++  0x00, 0xf1, 0x00, 0xf3, 0xf4, 0x00, 0xf6, 0xf7,
++  0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0x00, 0x00, 0xff,
++  0x00, 0x00, 0xc3, 0xe3, 0x00, 0x00, 0x00, 0x00,
++  0xfa, 0xfb
++};
++
++int
++main ()
++{
++  if (sizeof (a1) != 72
++      || sizeof (a2) != 64
++      || __builtin_memcmp (a1, a2, 64) != 0
++      || sizeof (a3) != 64
++      || __builtin_memcmp (a1, a3, 64) != 0
++      || sizeof (a4) != 65
++      || __builtin_memcmp (a1, a4, 65) != 0
++      || sizeof (a5) != 65
++      || __builtin_memcmp (a1, a5, 65) != 0
++      || sizeof (a6) != 66
++      || __builtin_memcmp (a1, a6, 66) != 0
++      || sizeof (a7) != 66
++      || __builtin_memcmp (a1, a7, 66) != 0)
++    __builtin_abort ();
++}
+-- 
+2.47.0
+