+++ /dev/null
->From a57cd35acf26ba6202ed6534a57f496464f431a1 Mon Sep 17 00:00:00 2001
-From: Szabolcs Nagy <nsz@port70.net>
-Date: Sat, 10 Aug 2019 23:14:40 +0000
-Subject: [PATCH] make relocation time symbol lookup and dlsym consistent
-
-Using common code path for all symbol lookups fixes three dlsym issues:
-
-- st_shndx of STT_TLS symbols were not checked and thus an undefined
- tls symbol reference could be incorrectly treated as a definition
- (the sysv hash lookup returns undefined symbols, gnu does not, so should
- be rare in practice).
-
-- symbol binding was not checked so a hidden symbol may be returned
- (in principle STB_LOCAL symbols may appear in the dynamic symbol table
- for hidden symbols, but linkers most likely don't produce it).
-
-- mips specific behaviour was not applied (ARCH_SYM_REJECT_UND) so
- undefined symbols may be returned on mips.
-
-always_inline is used to avoid relocation performance regression, the
-code generation for find_sym should not be affected.
----
- ldso/dynlink.c | 84 +++++++++++++++++++-------------------------------
- 1 file changed, 31 insertions(+), 53 deletions(-)
-
---- a/ldso/dynlink.c
-+++ b/ldso/dynlink.c
-@@ -283,12 +283,16 @@ static Sym *gnu_lookup_filtered(uint32_t
- #define ARCH_SYM_REJECT_UND(s) 0
- #endif
-
--static struct symdef find_sym(struct dso *dso, const char *s, int need_def)
-+#if defined(__GNUC__)
-+__attribute__((always_inline))
-+#endif
-+static inline struct symdef find_sym2(struct dso *dso, const char *s, int need_def, int use_deps)
- {
- uint32_t h = 0, gh = gnu_hash(s), gho = gh / (8*sizeof(size_t)), *ght;
- size_t ghm = 1ul << gh % (8*sizeof(size_t));
- struct symdef def = {0};
-- for (; dso; dso=dso->syms_next) {
-+ struct dso **deps = use_deps ? dso->deps : 0;
-+ for (; dso; dso=use_deps ? *deps++ : dso->syms_next) {
- Sym *sym;
- if ((ght = dso->ghashtab)) {
- sym = gnu_lookup_filtered(gh, ght, dso, s, gho, ghm);
-@@ -313,6 +317,11 @@ static struct symdef find_sym(struct dso
- return def;
- }
-
-+static struct symdef find_sym(struct dso *dso, const char *s, int need_def)
-+{
-+ return find_sym2(dso, s, need_def, 0);
-+}
-+
- static void do_relocs(struct dso *dso, size_t *rel, size_t rel_size, size_t stride)
- {
- unsigned char *base = dso->base;
-@@ -2118,58 +2127,27 @@ static void *addr2dso(size_t a)
-
- static void *do_dlsym(struct dso *p, const char *s, void *ra)
- {
-- size_t i;
-- uint32_t h = 0, gh = 0, *ght;
-- Sym *sym;
-- if (p == head || p == RTLD_DEFAULT || p == RTLD_NEXT) {
-- if (p == RTLD_DEFAULT) {
-- p = head;
-- } else if (p == RTLD_NEXT) {
-- p = addr2dso((size_t)ra);
-- if (!p) p=head;
-- p = p->next;
-- }
-- struct symdef def = find_sym(p, s, 0);
-- if (!def.sym) goto failed;
-- if ((def.sym->st_info&0xf) == STT_TLS)
-- return __tls_get_addr((tls_mod_off_t []){def.dso->tls_id, def.sym->st_value-DTP_OFFSET});
-- if (DL_FDPIC && (def.sym->st_info&0xf) == STT_FUNC)
-- return def.dso->funcdescs + (def.sym - def.dso->syms);
-- return laddr(def.dso, def.sym->st_value);
-- }
-- if (__dl_invalid_handle(p))
-+ int use_deps = 0;
-+ if (p == head || p == RTLD_DEFAULT) {
-+ p = head;
-+ } else if (p == RTLD_NEXT) {
-+ p = addr2dso((size_t)ra);
-+ if (!p) p=head;
-+ p = p->next;
-+ } else if (__dl_invalid_handle(p)) {
-+ return 0;
-+ } else
-+ use_deps = 1;
-+ struct symdef def = find_sym2(p, s, 0, use_deps);
-+ if (!def.sym) {
-+ error("Symbol not found: %s", s);
- return 0;
-- if ((ght = p->ghashtab)) {
-- gh = gnu_hash(s);
-- sym = gnu_lookup(gh, ght, p, s);
-- } else {
-- h = sysv_hash(s);
-- sym = sysv_lookup(s, h, p);
-- }
-- if (sym && (sym->st_info&0xf) == STT_TLS)
-- return __tls_get_addr((tls_mod_off_t []){p->tls_id, sym->st_value-DTP_OFFSET});
-- if (DL_FDPIC && sym && sym->st_shndx && (sym->st_info&0xf) == STT_FUNC)
-- return p->funcdescs + (sym - p->syms);
-- if (sym && sym->st_value && (1<<(sym->st_info&0xf) & OK_TYPES))
-- return laddr(p, sym->st_value);
-- for (i=0; p->deps[i]; i++) {
-- if ((ght = p->deps[i]->ghashtab)) {
-- if (!gh) gh = gnu_hash(s);
-- sym = gnu_lookup(gh, ght, p->deps[i], s);
-- } else {
-- if (!h) h = sysv_hash(s);
-- sym = sysv_lookup(s, h, p->deps[i]);
-- }
-- if (sym && (sym->st_info&0xf) == STT_TLS)
-- return __tls_get_addr((tls_mod_off_t []){p->deps[i]->tls_id, sym->st_value-DTP_OFFSET});
-- if (DL_FDPIC && sym && sym->st_shndx && (sym->st_info&0xf) == STT_FUNC)
-- return p->deps[i]->funcdescs + (sym - p->deps[i]->syms);
-- if (sym && sym->st_value && (1<<(sym->st_info&0xf) & OK_TYPES))
-- return laddr(p->deps[i], sym->st_value);
- }
--failed:
-- error("Symbol not found: %s", s);
-- return 0;
-+ if ((def.sym->st_info&0xf) == STT_TLS)
-+ return __tls_get_addr((tls_mod_off_t []){def.dso->tls_id, def.sym->st_value-DTP_OFFSET});
-+ if (DL_FDPIC && (def.sym->st_info&0xf) == STT_FUNC)
-+ return def.dso->funcdescs + (def.sym - def.dso->syms);
-+ return laddr(def.dso, def.sym->st_value);
- }
-
- int dladdr(const void *addr_arg, Dl_info *info)
+++ /dev/null
-From f3ed8bfe8a82af1870ddc8696ed4cc1d5aa6b441 Mon Sep 17 00:00:00 2001
-From: Rich Felker <dalias@aerifal.cx>
-Date: Mon, 5 Aug 2019 18:41:47 -0400
-Subject: fix x87 stack imbalance in corner cases of i386 math asm
-
-commit 31c5fb80b9eae86f801be4f46025bc6532a554c5 introduced underflow
-code paths for the i386 math asm, along with checks on the fpu status
-word to skip the underflow-generation instructions if the underflow
-flag was already raised. unfortunately, at least one such path, in
-log1p, returned with 2 items on the x87 stack rather than just 1 item
-for the return value. this is a violation of the ABI's calling
-convention, and could cause subsequent floating point code to produce
-NANs due to x87 stack overflow. if floating point results are used in
-flow control, this can lead to runaway wrong code execution.
-
-rather than reviewing each "underflow already raised" code path for
-correctness, remove them all. they're likely slower than just
-performing the underflow code unconditionally, and significantly more
-complex.
-
-all of this code should be ripped out and replaced by C source files
-with inline asm. doing so would preclude this kind of error by having
-the compiler perform all x87 stack register allocation and stack
-manipulation, and would produce comparable or better code. however
-such a change is a much larger project.
----
- src/math/i386/asin.s | 10 ++--------
- src/math/i386/atan.s | 7 ++-----
- src/math/i386/atan2.s | 5 +----
- src/math/i386/atan2f.s | 5 +----
- src/math/i386/atanf.s | 7 ++-----
- src/math/i386/exp.s | 10 ++--------
- src/math/i386/log1p.s | 7 ++-----
- src/math/i386/log1pf.s | 7 ++-----
- 8 files changed, 14 insertions(+), 44 deletions(-)
-
---- a/src/math/i386/asin.s
-+++ b/src/math/i386/asin.s
-@@ -7,13 +7,10 @@ asinf:
- cmp $0x01000000,%eax
- jae 1f
- # subnormal x, return x with underflow
-- fnstsw %ax
-- and $16,%ax
-- jnz 2f
- fld %st(0)
- fmul %st(1)
- fstps 4(%esp)
--2: ret
-+ ret
-
- .global asinl
- .type asinl,@function
-@@ -30,11 +27,8 @@ asin:
- cmp $0x00200000,%eax
- jae 1f
- # subnormal x, return x with underflow
-- fnstsw %ax
-- and $16,%ax
-- jnz 2f
- fsts 4(%esp)
--2: ret
-+ ret
- 1: fld %st(0)
- fld1
- fsub %st(0),%st(1)
---- a/src/math/i386/atan.s
-+++ b/src/math/i386/atan.s
-@@ -10,8 +10,5 @@ atan:
- fpatan
- ret
- # subnormal x, return x with underflow
--1: fnstsw %ax
-- and $16,%ax
-- jnz 2f
-- fsts 4(%esp)
--2: ret
-+1: fsts 4(%esp)
-+ ret
---- a/src/math/i386/atan2.s
-+++ b/src/math/i386/atan2.s
-@@ -10,8 +10,5 @@ atan2:
- cmp $0x00200000,%eax
- jae 1f
- # subnormal x, return x with underflow
-- fnstsw %ax
-- and $16,%ax
-- jnz 1f
- fsts 4(%esp)
--1: ret
-+ ret
---- a/src/math/i386/atan2f.s
-+++ b/src/math/i386/atan2f.s
-@@ -10,10 +10,7 @@ atan2f:
- cmp $0x01000000,%eax
- jae 1f
- # subnormal x, return x with underflow
-- fnstsw %ax
-- and $16,%ax
-- jnz 1f
- fld %st(0)
- fmul %st(1)
- fstps 4(%esp)
--1: ret
-+ ret
---- a/src/math/i386/atanf.s
-+++ b/src/math/i386/atanf.s
-@@ -10,10 +10,7 @@ atanf:
- fpatan
- ret
- # subnormal x, return x with underflow
--1: fnstsw %ax
-- and $16,%ax
-- jnz 2f
-- fld %st(0)
-+1: fld %st(0)
- fmul %st(1)
- fstps 4(%esp)
--2: ret
-+ ret
---- a/src/math/i386/exp.s
-+++ b/src/math/i386/exp.s
-@@ -7,13 +7,10 @@ expm1f:
- cmp $0x01000000,%eax
- jae 1f
- # subnormal x, return x with underflow
-- fnstsw %ax
-- and $16,%ax
-- jnz 2f
- fld %st(0)
- fmul %st(1)
- fstps 4(%esp)
--2: ret
-+ ret
-
- .global expm1l
- .type expm1l,@function
-@@ -30,11 +27,8 @@ expm1:
- cmp $0x00200000,%eax
- jae 1f
- # subnormal x, return x with underflow
-- fnstsw %ax
-- and $16,%ax
-- jnz 2f
- fsts 4(%esp)
--2: ret
-+ ret
- 1: fldl2e
- fmulp
- mov $0xc2820000,%eax
---- a/src/math/i386/log1p.s
-+++ b/src/math/i386/log1p.s
-@@ -16,9 +16,6 @@ log1p:
- fyl2x
- ret
- # subnormal x, return x with underflow
--2: fnstsw %ax
-- and $16,%ax
-- jnz 1f
-- fsts 4(%esp)
-+2: fsts 4(%esp)
- fstp %st(1)
--1: ret
-+ ret
---- a/src/math/i386/log1pf.s
-+++ b/src/math/i386/log1pf.s
-@@ -16,10 +16,7 @@ log1pf:
- fyl2x
- ret
- # subnormal x, return x with underflow
--2: fnstsw %ax
-- and $16,%ax
-- jnz 1f
-- fxch
-+2: fxch
- fmul %st(1)
- fstps 4(%esp)
--1: ret
-+ ret
projects / oweals / openwrt.git / commitdiff