Jakub Bogusz from pld-linux dot org writes:

Hello,

I managed to get ldso (and thus shared linking to uClibc) to work on
sparc (actually sparc64 kernel with 32-bit userspace), at least on
simple "hello world" program (more complex ones not tested).

Some notes on attached patch (against 0.9.26, would require some work
to apply on current CVS - but I tested 0.9.26, not CVS):

- ELF magic cannot be examined by _dl_strncmp so early, probably because of
  string constant, like on ppc/mips/sh
  (note that early SEND_STDERR still crashes when trying to do _dl_strlen
   - I suppose that string constants require relocation; but adding
   load_addr didn't help, just ELF header was displayed instead of crash)
- mmap() is syscall6 like on ppc/mips/sh, not old i386 mmap()
- for generic sparc (i.e. not sparcv8/sparcv9) gcc produces .udiv/.urem
  calls for unsigned integer / and % operators - so these operations
  must be avoided. I copied do_rem definition from arm header.
  But / and % are used also in _dl_simple_ltoa() and
  _dl_simple_ltoahex(); in ltoahex gcc optimizes it to shifts (but
  I think it's safer to use shifts explicitly, not rely on
  optimization...).
  I changed % in ltoa to do_rem, but as there was no do_div definition,
  I changed all "%d" specifiers to "%x" to avoid crashes (this changes
  wouldn't be needed if _dl_simple_ltoa() were fixed to not use
  division on sparc).
- "#define SOLARIS_COMPATIBLE" in ld_sysdep.h broke ldso on Linux
  because of redefining _dl_linux_resolve only in some places (it was
  still referenced in INIT_GOT before redefinition). So
  _dl_linux_resolve redefinition should be moved before INIT_GOT
  definition or removed.
- sparc64 kernel requires mmap() addresses to be aligned to 8192, not
  4096, otherwise mmap() call failed
- reloc_entry must be shifted by 10, not 12 (I found similar operation
  in glibc sources)

Aside of sparc-specific fixes:
- I moved some _dl_dprintf()s inside if(_dl_debug_*) conditions (to avoid
  debugging messages when LD_DEBUG is not defined)
- it seems that there was possible off-by-one in ltoa and ltoahex?
  they are called with char[22] as 1st argument, and then '\0' is stored
  in local[22] (_before_ p decrementation)... or am I missing something?
  If not, fix is included in patch.

6 files changed, 92 insertions, 16 deletions

diff --git a/ldso/include/dl-string.h b/ldso/include/dl-string.h
index cbb867010..05f8d7d32 100644
--- a/ldso/include/dl-string.h
+++ b/ldso/include/dl-string.h
@@ -204,15 +204,18 @@ static inline char *_dl_get_last_path_component(char *path)
 }
 
 /* Early on, we can't call printf, so use this to print out
- * numbers using the SEND_STDERR() macro */
+ * numbers using the SEND_STDERR() macro.  Avoid using mod
+ * or using long division */
 static inline char *_dl_simple_ltoa(char * local, unsigned long i)
 {
 	/* 21 digits plus null terminator, good for 64-bit or smaller ints */
-	char *p = &local[22];
+	char *p = &local[21];
 	*p-- = '\0';
 	do {
-		*p-- = '0' + i % 10;
-		i /= 10;
+	    char temp;
+	    do_rem(temp, i, 10);
+	    *p-- = '0' + temp;
+	    i /= 10;
 	} while (i > 0);
 	return p + 1;
 }
@@ -220,15 +223,15 @@ static inline char *_dl_simple_ltoa(char * local, unsigned long i)
 static inline char *_dl_simple_ltoahex(char * local, unsigned long i)
 {
 	/* 21 digits plus null terminator, good for 64-bit or smaller ints */
-	char *p = &local[22];
+	char *p = &local[21];
 	*p-- = '\0';
 	do {
-		char temp = i % 0x10;
+		char temp = i & 0xf;
 		if (temp <= 0x09)
 		    *p-- = '0' + temp;
 		else
 		    *p-- = 'a' - 0x0a + temp;
-		i /= 0x10;
+		i >>= 4;
 	} while (i > 0);
 	*p-- = 'x';
 	*p-- = '0';
diff --git a/ldso/include/dl-syscall.h b/ldso/include/dl-syscall.h
index 184c12fb1..fbe852d18 100644
--- a/ldso/include/dl-syscall.h
+++ b/ldso/include/dl-syscall.h
@@ -104,7 +104,7 @@ static inline _syscall0(gid_t, _dl_getpid);
 static inline _syscall3(int, _dl_readlink, const char *, path, char *, buf, size_t, bufsiz);
 
 #ifdef __NR_mmap
-#if defined(__powerpc__) || defined(__mips__) || defined(__sh__)
+#if defined(__powerpc__) || defined(__mips__) || defined(__sh__) || defined(__sparc__)
 /* PowerPC, MIPS and SuperH have a different calling convention for mmap(). */
 #define __NR__dl_mmap __NR_mmap
 static inline _syscall6(void *, _dl_mmap, void *, start, size_t, length,
diff --git a/ldso/ldso/dl-elf.c b/ldso/ldso/dl-elf.c
index dc20252d0..d27e4ccb3 100644
--- a/ldso/ldso/dl-elf.c
+++ b/ldso/ldso/dl-elf.c
@@ -720,7 +720,7 @@ struct elf_resolve *_dl_load_elf_shared_library(int secure,
 		_dl_dprintf(2, "\n\tfile='%s';  generating link map\n", libname);
 		_dl_dprintf(2, "\t\tdynamic: %x  base: %x   size: %x\n",
 				dynamic_addr, libaddr, dynamic_size);
-		_dl_dprintf(2, "\t\t  entry: %x  phdr: %x  phnum: %d\n\n",
+		_dl_dprintf(2, "\t\t  entry: %x  phdr: %x  phnum: %x\n\n",
 				epnt->e_entry + libaddr, tpnt->ppnt, tpnt->n_phent);
 
 	}
diff --git a/ldso/ldso/sparc/dl-syscalls.h b/ldso/ldso/sparc/dl-syscalls.h
index 218e8a289..cf21061fd 100644
--- a/ldso/ldso/sparc/dl-syscalls.h
+++ b/ldso/ldso/sparc/dl-syscalls.h
@@ -158,3 +158,29 @@ if (__res < -255 || __res>=0) \
 _dl_errno = -__res; \
 return -1; \
 }
+
+#define _syscall6(type,name,type1,arg1,type2,arg2,type3,arg3,type4,arg4, \
+	  type5,arg5,type6,arg6) \
+type name (type1 arg1,type2 arg2,type3 arg3,type4 arg4,type5 arg5,type6 arg6) \
+{ \
+long __res; \
+register long __g1 __asm__ ("g1") = __NR_##name; \
+register long __o0 __asm__ ("o0") = (long)(arg1); \
+register long __o1 __asm__ ("o1") = (long)(arg2); \
+register long __o2 __asm__ ("o2") = (long)(arg3); \
+register long __o3 __asm__ ("o3") = (long)(arg4); \
+register long __o4 __asm__ ("o4") = (long)(arg5); \
+register long __o5 __asm__ ("o5") = (long)(arg6); \
+__asm__ __volatile__ ("t 0x10\n\t" \
+		      "bcc 1f\n\t" \
+		      "mov %%o0, %0\n\t" \
+		      "sub %%g0, %%o0, %0\n\t" \
+		      "1:\n\t" \
+		      : "=r" (__res), "=&r" (__o0) \
+		      : "1" (__o0), "r" (__o1), "r" (__o2), "r" (__o3), "r" (__o4), "r" (__o5), "r" (__g1) \
+		      : "cc"); \
+if (__res < -255 || __res>=0) \
+	return (type) __res; \
+/*errno = -__res; */\
+return -1; \
+}
diff --git a/ldso/ldso/sparc/dl-sysdep.h b/ldso/ldso/sparc/dl-sysdep.h
index c6ac89e11..ddf74e21e 100644
--- a/ldso/ldso/sparc/dl-sysdep.h
+++ b/ldso/ldso/sparc/dl-sysdep.h
@@ -41,9 +41,51 @@ extern unsigned int _dl_linux_resolver(unsigned int reloc_entry,
 /*
  * Define this if you want a dynamic loader that works on Solaris.
  */
+#ifndef __linux__
 #define SOLARIS_COMPATIBLE
+#endif
+
+#ifndef COMPILE_ASM
+/* Cheap modulo implementation, taken from arm/ld_sysdep.h. */
+static inline unsigned long
+sparc_mod(unsigned long m, unsigned long p)
+{
+	unsigned long i, t, inc;
+
+	i = p;
+	t = 0;
+
+	while (!(i & (1 << 31))) {
+		i <<= 1;
+		t++;
+	}
+
+	t--;
+
+	for (inc = t; inc > 2; inc--) {
+		i = p << inc;
 
-#define do_rem(result, n, base)	    result = (n % base)
+		if (i & (1 << 31))
+			break;
+
+		while (m >= i) {
+			m -= i;
+			i <<= 1;
+			if (i & (1 << 31))
+				break;
+			if (i < p)
+				break;
+		}
+	}
+
+	while (m >= p)
+		m -= p;
+
+	return m;
+}
+
+#define do_rem(result, n, base) result = sparc_mod(n, base);
+#endif
 
 /*
  * dbx wants the binder to have a specific name.  Mustn't disappoint it.
@@ -53,9 +95,10 @@ extern unsigned int _dl_linux_resolver(unsigned int reloc_entry,
 #endif
 
 /* 4096 bytes alignment */
-#define PAGE_ALIGN 0xfffff000
-#define ADDR_ALIGN 0xfff
-#define OFFS_ALIGN 0x7ffff000
+/* ...but 8192 is required for mmap() on sparc64 kernel */
+#define PAGE_ALIGN 0xffffe000
+#define ADDR_ALIGN 0x1fff
+#define OFFS_ALIGN 0x7fffe000
 
 /* ELF_RTYPE_CLASS_PLT iff TYPE describes relocation of a PLT entry, so
    PLT entries should not be allowed to define the value.
diff --git a/ldso/ldso/sparc/elfinterp.c b/ldso/ldso/sparc/elfinterp.c
index 41bde4fa8..98435cbd4 100644
--- a/ldso/ldso/sparc/elfinterp.c
+++ b/ldso/ldso/sparc/elfinterp.c
@@ -72,7 +72,7 @@ unsigned int _dl_linux_resolver(unsigned int reloc_entry, unsigned int * plt)
   /*
    * Generate the correct relocation index into the .rela.plt section.
    */
-  reloc_entry = (reloc_entry >> 12) - 0xc;
+  reloc_entry = (reloc_entry >> 10) - 0xc;
 
   this_reloc = (Elf32_Rela *) ((char *) rel_addr + reloc_entry);
 
@@ -82,10 +82,14 @@ unsigned int _dl_linux_resolver(unsigned int reloc_entry, unsigned int * plt)
   symtab =  (Elf32_Sym *) (tpnt->dynamic_info[DT_SYMTAB] + tpnt->loadaddr);
   strtab = (char *) (tpnt->dynamic_info[DT_STRTAB] + tpnt->loadaddr);
 
+#ifdef __SUPPORT_LD_DEBUG__
+  if (_dl_debug_symbols) {
   _dl_dprintf(2, "tpnt = %x\n", tpnt);
   _dl_dprintf(2, "reloc = %x\n", this_reloc);
   _dl_dprintf(2, "symtab = %x\n", symtab);
   _dl_dprintf(2, "strtab = %x\n", strtab);
+  }
+#endif
 
 
   if (unlikely(reloc_type != R_SPARC_JMP_SLOT)) {
@@ -98,10 +102,10 @@ unsigned int _dl_linux_resolver(unsigned int reloc_entry, unsigned int * plt)
   instr_addr  = ((int)this_reloc->r_offset  + (int)tpnt->loadaddr);
   got_addr = (char **) instr_addr;
 
-  _dl_dprintf(2, "symtab_index %d\n", symtab_index);
-
 #ifdef __SUPPORT_LD_DEBUG__
   if (_dl_debug_symbols) {
+  _dl_dprintf(2, "symtab_index %x\n", symtab_index);
+
 	  _dl_dprintf(2, "Resolving symbol %s\n",
 			  strtab + symtab[symtab_index].st_name);
   }