From 302e24671942051d70707586cf8c605a5815edac Mon Sep 17 00:00:00 2001
From: lemon <lsof@mailbox.org>
Date: Mon, 15 Dec 2025 22:39:52 +0100
Subject: create distinct interned string type

Interned strings are used pervasively, so it's a good idea to add a
layer of type safety to differentiate them from general cstrs and avoid
potential bugs from comparing non-interned and interned strings. Not
that that's happened so far that I can remember, but it could.

I'm 90% sure it's legal to alias `struct {char c;}` pointers with `char`
pointers. This specific typedef gives type safety but with a simple
one-way `internstr -> const char *` typecast (with `&istr->c`).
Converting the other way around is more intentional: a straight up cast
`(internstr)cstr` which sticks out as unchecked and probably wrong, or
calling the intern(cstr) function, which is the right way.
---
 obj/elf.c | 14 +++++++-------
 obj/obj.c | 14 +++++++-------
 obj/obj.h |  8 ++++----
 3 files changed, 18 insertions(+), 18 deletions(-)

(limited to 'obj')

diff --git a/obj/elf.c b/obj/elf.c
index 521a159..ea326b6 100644
--- a/obj/elf.c
+++ b/obj/elf.c
@@ -29,7 +29,7 @@ struct reloc {
    vlong addend;
    union {
       uint symidx;
-      const char *symname;
+      internstr symname;
    };
 };
 static vec_of(struct reloc) relocs;
@@ -78,7 +78,7 @@ str2idx(const char *s)
 }
 
 static struct sym *
-findsym(const char *s)
+findsym(internstr s)
 {
    ushort *idx = pmap_get(&symht, s);
    return idx ? &symtab.p[*idx] : NULL;
@@ -102,7 +102,7 @@ static const char sect2ndx[] = {
 };
 
 enum section
-elfhassym(const char *nam, uint *value)
+elfhassym(internstr nam, uint *value)
 {
    struct sym *sym = findsym(nam);
    if (sym) {
@@ -113,12 +113,12 @@ elfhassym(const char *nam, uint *value)
 }
 
 void
-elfaddsym(const char *nam, int info, enum section sect, uvlong value, uvlong size)
+elfaddsym(internstr nam, int info, enum section sect, uvlong value, uvlong size)
 {
    struct sym *sym = findsym(nam), sym0;
    if (!sym) {
       sym = &sym0;
-      sym->name = str2idx(nam);
+      sym->name = str2idx(&nam->c);
    }
    sym->bind = info >> 4;
    sym->type = info & 0xF;
@@ -145,7 +145,7 @@ static const ushort relktab[][NRELOCKIND] = {
 };
 
 void
-elfreloc(const char *sym, enum relockind kind, enum section section, uint off, vlong addend)
+elfreloc(internstr sym, enum relockind kind, enum section section, uint off, vlong addend)
 {
    switch (section) {
    default: assert(0);
@@ -401,7 +401,7 @@ elffini(struct wbuf *out)
          rel->symidx = idx < ndefsym ? defsym2idx[idx] : idx;
       } else {
          assert(symtab.n < 1<<16);
-         vpush(&symtab, ((struct sym) { str2idx(rel->symname), .bind = STB_GLOBAL, .type = STT_NOTYPE, .shndx = SHN_UND }));
+         vpush(&symtab, ((struct sym) { str2idx(&rel->symname->c), .bind = STB_GLOBAL, .type = STT_NOTYPE, .shndx = SHN_UND }));
          pmap_set(&symht, rel->symname, symtab.n-1);
          rel->symidx = symtab.n-1;
       }
diff --git a/obj/obj.c b/obj/obj.c
index 2d8b497..12f0db7 100644
--- a/obj/obj.c
+++ b/obj/obj.c
@@ -6,9 +6,9 @@
 
 
 void elfinit(void);
-enum section elfhassym(const char *, uint *value);
-void elfaddsym(const char *, int info, enum section, uvlong value, uvlong size);
-void elfreloc(const char *sym, enum relockind, enum section, uint off, vlong addend);
+enum section elfhassym(internstr , uint *value);
+void elfaddsym(internstr , int info, enum section, uvlong value, uvlong size);
+void elfreloc(internstr sym, enum relockind, enum section, uint off, vlong addend);
 void elffini(struct wbuf *);
 
 struct objfile objout;
@@ -30,7 +30,7 @@ objini(const char *infile, const char *outfile)
 }
 
 void
-objdeffunc(const char *nam, bool globl, uint off, uint siz)
+objdeffunc(internstr nam, bool globl, uint off, uint siz)
 {
    switch (mctarg->objkind) {
    case OBJELF:
@@ -40,13 +40,13 @@ objdeffunc(const char *nam, bool globl, uint off, uint siz)
 }
 
 enum section
-objhassym(const char *name, uint *off)
+objhassym(internstr name, uint *off)
 {
    return elfhassym(name, off);
 }
 
 uint
-objnewdat(const char *name, enum section sec, bool globl, uint siz, uint align)
+objnewdat(internstr name, enum section sec, bool globl, uint siz, uint align)
 {
    struct objfile *o = &objout;
    uint off;
@@ -90,7 +90,7 @@ objnewdat(const char *name, enum section sec, bool globl, uint siz, uint align)
 }
 
 void
-objreloc(const char *sym, enum relockind reloc, enum section section, uint off, vlong addend)
+objreloc(internstr sym, enum relockind reloc, enum section section, uint off, vlong addend)
 {
    switch (mctarg->objkind) {
    case OBJELF:
diff --git a/obj/obj.h b/obj/obj.h
index 59769ff..77da99a 100644
--- a/obj/obj.h
+++ b/obj/obj.h
@@ -22,10 +22,10 @@ enum relockind {
 enum section { Snone, Stext, Srodata, Sdata, Sbss };
 
 void objini(const char *infile, const char *outfile);
-void objdeffunc(const char *nam, bool globl, uint off, uint siz);
-enum section objhassym(const char *name, uint *off);
-uint objnewdat(const char *name, enum section, bool globl, uint siz, uint align);
-void objreloc(const char *sym, enum relockind, enum section, uint off, vlong addend);
+void objdeffunc(internstr nam, bool globl, uint off, uint siz);
+enum section objhassym(internstr name, uint *off);
+uint objnewdat(internstr name, enum section, bool globl, uint siz, uint align);
+void objreloc(internstr sym, enum relockind, enum section, uint off, vlong addend);
 void objfini(void);
 
 /* vim:set ts=3 sw=3 expandtab: */
-- 
cgit v1.2.3