Merge pull request #95 from sysprog21/style

Style tweaks

Author: jserv

src/core/bootstrap.c

@@ -57,7 +57,9 @@ static bool append_boot_region(mem_region_t *regions,
 }
 
 /* Emit one mem_region_t per PT_LOAD segment of an ELF image, offset by the
- * caller-supplied load base. Returns false if the boot region array fills up.
+ * caller-supplied load base.
+ *
+ * Returns false if the boot region array fills up.
  */
 static bool append_elf_segment_regions(mem_region_t *regions,
                                        int *nregions,
@@ -77,8 +79,8 @@ static bool append_elf_segment_regions(mem_region_t *regions,
 
 /* Register one semantic guest_region_t per PT_LOAD segment of an ELF image.
  * va_load_base controls the guest-visible range, gpa_load_base controls the
- * backing GPA recorded in region metadata, and path is used for
- * /proc/self/maps reporting.
+ * backing GPA recorded in region metadata, and path is used for /proc/self/maps
+ * reporting.
  */
 static void register_elf_segment_regions(guest_t *g,
                                          const elf_info_t *info,
@@ -519,10 +521,10 @@ int guest_bootstrap_prepare(guest_t *g,
 
     t0 = startup_trace_now_ns();
     if (want_rosetta) {
-        /* /proc/self/maps for a rosetta guest reports the rosetta translator
-         * as a single anonymous region covering [VA, VA+size). The original
-         * x86_64 binary is not loaded into guest memory; rosetta exposes it
-         * via fd 3 once rosetta_finalize pre-opens it.
+        /* /proc/self/maps for a rosetta guest reports the rosetta translator as
+         * a single anonymous region covering [VA, VA+size). The original x86_64
+         * binary is not loaded into guest memory; rosetta exposes it via fd 3
+         * once rosetta_finalize pre-opens it.
          */
         register_elf_segment_regions(g, &rr.rosetta_info, 0,
                                      g->rosetta_guest_base - g->rosetta_va_base,
@@ -612,10 +614,10 @@ int guest_bootstrap_prepare(guest_t *g,
         return -1;
     }
     startup_trace_step("build_linux_stack", t0);
-    /* rosetta_argv was copied into the guest stack; the host allocation is
-     * no longer needed. The strings themselves are constants (ROSETTA_PATH)
-     * or owned by the caller (binary_path, guest_argv entries) so freeing
-     * just the array is safe.
+    /* rosetta_argv was copied into the guest stack; the host allocation is no
+     * longer needed. The strings themselves are constants (ROSETTA_PATH) or
+     * owned by the caller (binary_path, guest_argv entries) so freeing just the
+     * array is safe.
      */
     free(rosetta_argv);
 
@@ -650,8 +652,8 @@ int guest_bootstrap_create_vcpu(guest_t *g,
     uint64_t t0;
     /* Rosetta needs TTBR1 walks enabled and TBI1=1 so the kbuf window at
      * KBUF_VA_BASE (bits-63-set) resolves and TaggedPointer extraction keeps
-     * working. Aarch64 guests stay on the EPD1=1 variant which keeps the
-     * upper VA range fault-clean.
+     * working. Aarch64 guests stay on the EPD1=1 variant which keeps the upper
+     * VA range fault-clean.
      */
     uint64_t tcr_value = g->is_rosetta ? TCR_EL1_VALUE_KBUF : TCR_EL1_VALUE;
     uint64_t ttbr1_value = g->is_rosetta ? g->ttbr1 : 0;
@@ -684,10 +686,10 @@ int guest_bootstrap_create_vcpu(guest_t *g,
     HV_CHECK(hv_vcpu_set_sys_reg(vcpu, HV_SYS_REG_SP_EL0, sp_ipa));
     HV_CHECK(hv_vcpu_set_sys_reg(vcpu, HV_SYS_REG_SP_EL1, el1_sp));
 
-    /* Round-trip a sentinel through TPIDR_EL1 before installing the real
-     * value. Validates only the hv_vcpu_{set,get}_sys_reg pre-run round
-     * trip, not preservation across hv_vcpu_run -- the test-shim-identity
-     * microbench is the end-to-end check for that.
+    /* Round-trip a sentinel through TPIDR_EL1 before installing the real value.
+     * Validates only the hv_vcpu_{set,get}_sys_reg pre-run round trip, not
+     * preservation across hv_vcpu_run -- the test-shim-identity microbench is
+     * the end-to-end check for that.
      */
     if (shim_globals_self_test(vcpu) < 0)
         return -1;
@@ -790,11 +792,11 @@ int guest_bootstrap_rosetta_post_reset(guest_t *g,
         !out_stack_pointer || !out_ttbr0)
         return -1;
 
-    /* Re-anchor brk/stack to the Rosetta defaults. guest_reset already
-     * restored mmap_next/mmap_end/mmap_rx_* to their initial values, but
-     * brk/stack were tuned for the previous image, so reset them here.
-     * The x86_64 target binary lives behind fd 3, not in guest memory,
-     * so brk_base does not move with the target's load_max.
+    /* Re-anchor brk/stack to the Rosetta defaults. guest_reset already restored
+     * mmap_next/mmap_end/mmap_rx_* to their initial values, but brk/stack were
+     * tuned for the previous image, so reset them here. The x86_64 target
+     * binary lives behind fd 3, not in guest memory, so brk_base does not move
+     * with the target's load_max.
      */
     g->elf_load_min = ELF_DEFAULT_BASE;
     g->brk_base = BRK_BASE_DEFAULT;

src/core/bootstrap.h

@@ -60,23 +60,24 @@ int guest_bootstrap_create_vcpu(guest_t *g,
  *     proc_set_rosetta_active(true) so rosetta_prepare and rosettad gates
  *     see the right runtime state
  *
- * elf_host_path is the macOS filesystem path used by rosetta_prepare to
- * open the binary (after sysroot/FUSE resolution). elf_guest_path is the
- * unresolved guest-visible path published through proc_set_elf_path and
- * rosetta_finalize's /proc/self/cmdline rewrite.
+ * elf_host_path is the macOS filesystem path used by rosetta_prepare to open
+ * the binary (after sysroot/FUSE resolution). elf_guest_path is the unresolved
+ * guest-visible path published through proc_set_elf_path and rosetta_finalize's
+ * /proc/self/cmdline rewrite.
  *
- * The helper runs rosetta_prepare, appends every region the page-table
- * builder needs, rebuilds page tables, registers guest_region_t entries
- * for /proc/self/maps, runs rosetta_finalize (pre-opens fd 3, installs the
- * kbuf user alias, publishes the binfmt-misc argv via proc_set_cmdline),
- * and builds the initial Linux stack using the rosetta image as the
- * AT_PHDR/AT_BASE ELF metadata. It does NOT touch the vCPU sysregs --
- * the caller writes TCR_EL1, TTBR0_EL1, TTBR1_EL1, ELR_EL1, SP_EL0, and
- * PC itself once the out_* fields are returned.
+ * The helper runs rosetta_prepare, appends every region the page-table builder
+ * needs, rebuilds page tables, registers guest_region_t entries for
+ * /proc/self/maps, runs rosetta_finalize (pre-opens fd 3, installs the kbuf
+ * user alias, publishes the binfmt-misc argv via proc_set_cmdline), and builds
+ * the initial Linux stack using the rosetta image as the AT_PHDR/AT_BASE ELF
+ * metadata. It does NOT touch the vCPU sysregs -- the caller writes TCR_EL1,
+ * TTBR0_EL1, TTBR1_EL1, ELR_EL1, SP_EL0, and PC itself once the out_* fields
+ * are returned.
  *
- * Returns 0 on success with out_entry_point, out_stack_pointer, out_ttbr0
- * set. Returns -1 on any internal failure; the caller is past the point of
- * no return and treats that as fatal.
+ * Returns 0 on success with out_entry_point, out_stack_pointer, out_ttbr0 set.
+ *
+ * Returns -1 on any internal failure; the caller is past the point of no return
+ * and treats that as fatal.
  */
 int guest_bootstrap_rosetta_post_reset(guest_t *g,
                                        const char *elf_host_path,

src/core/elf.c

@@ -145,8 +145,8 @@ int elf_load(const char *path, elf_info_t *info)
                 long saved_pos = ftell(f);
                 if (fseek(f, (long) ph->p_offset, SEEK_SET) == 0) {
                     size_t n = fread(info->interp_path, 1, interp_len, f);
-                    /* interp_len includes the NUL from the ELF file.
-                     * On short read, clear the path (unusable). On full read,
+                    /* interp_len includes the NUL from the ELF file. On short
+                     * read, clear the path (unusable). On full read,
                      * force-terminate as insurance.
                      */
                     if (n < interp_len)
@@ -193,9 +193,9 @@ int elf_load(const char *path, elf_info_t *info)
         return -1;
     }
 
-    /* Store program header file offset for later phdr_gpa calculation.
-     * The loader places program headers at the same GPA as they would be in
-     * the first PT_LOAD segment (they are typically within it).
+    /* Store program header file offset for later phdr_gpa calculation. The
+     * loader places program headers at the same GPA as they would be in the
+     * first PT_LOAD segment (they are typically within it).
      */
     info->phdr_gpa = info->load_min + ehdr.e_phoff;
 
@@ -213,9 +213,9 @@ int elf_map_segments(const elf_info_t *info,
                      uint64_t infra_hi)
 {
     /* Half-open intersection test for [a, a+alen) and [b, b+blen). When
-     * infra_lo == infra_hi the caller opted out (early bring-up before
-     * guest_t is wired up); the host-side writes that follow still get
-     * the existing guest_size bound check.
+     * infra_lo == infra_hi the caller opted out (early bring-up before guest_t
+     * is wired up); the host-side writes that follow still get the existing
+     * guest_size bound check.
      */
     bool infra_active = infra_lo < infra_hi;
     FILE *f = fopen(path, "rb");
@@ -231,9 +231,9 @@ int elf_map_segments(const elf_info_t *info,
         return -1;
     }
 
-    /* Read and parse program headers again to get file offsets. The size
-     * was already bound-checked during elf_load(); recheck defensively in
-     * case the header sizes changed since (e.g. corrupt file races).
+    /* Read and parse program headers again to get file offsets. The size was
+     * already bound-checked during elf_load(); recheck defensively in case the
+     * header sizes changed since (e.g. corrupt file races).
      */
     size_t ph_total = (size_t) ehdr.e_phnum * ehdr.e_phentsize;
     if (ph_total == 0 || ph_total > 65536) {
@@ -327,28 +327,28 @@ int elf_map_segments(const elf_info_t *info,
             return -1;
         }
 
-        /* PT_LOAD with memsz == 0 maps no bytes, but the page-tail zero
-         * extent below still rounds up to the next page boundary. For an
-         * unaligned gpa that means a crafted ELF could splat zeros across
-         * the tail of a previously loaded segment in the same page, or
-         * trip the infra-overlap check with no live mapping behind it.
-         * Linux ignores zero-memsz PT_LOADs; mirror that here.
+        /* PT_LOAD with memsz == 0 maps no bytes, but the page-tail zero extent
+         * below still rounds up to the next page boundary. For an unaligned gpa
+         * that means a crafted ELF could splat zeros across the tail of a
+         * previously loaded segment in the same page, or trip the infra-overlap
+         * check with no live mapping behind it. Linux ignores zero-memsz
+         * PT_LOADs; mirror that here.
          */
         if (memsz == 0) {
             seg_idx++;
             continue;
         }
 
-        /* The host memset zeros up to the next page boundary AFTER the
-         * segment ends, so the infra-overlap check has to use the same
-         * rounded extent. The end is PAGE_ALIGN_UP(gpa + memsz) rather
-         * than gpa + PAGE_ALIGN_UP(memsz) because gpa is not always
-         * page-aligned (e.g. ld.so's RW segment at vaddr 0x2f650): with
-         * the older bytes-from-gpa formula the page covering the last
-         * memsz byte kept its mid-page tail untouched, and execve into a
-         * dynamic-linked target then read stale state from the prior
-         * incarnation of the same interpreter at offsets ld.so allocates
-         * from beyond memsz (e.g. the first link_map in _dl_new_object).
+        /* The host memset zeros up to the next page boundary AFTER the segment
+         * ends, so the infra-overlap check has to use the same rounded extent.
+         * The end is PAGE_ALIGN_UP(gpa + memsz) rather than gpa +
+         * PAGE_ALIGN_UP(memsz) because gpa is not always page-aligned (e.g.
+         * ld.so's RW segment at vaddr 0x2f650): with the older bytes-from-gpa
+         * formula the page covering the last memsz byte kept its mid-page tail
+         * untouched, and execve into a dynamic-linked target then read stale
+         * state from the prior incarnation of the same interpreter at offsets
+         * ld.so allocates from beyond memsz (e.g. the first link_map in
+         * _dl_new_object).
          */
         uint64_t zero_len = PAGE_ALIGN_UP(gpa + memsz) - gpa;
         if (gpa + zero_len > guest_size)
@@ -415,8 +415,8 @@ void elf_resolve_interp(const char *sysroot,
         if (access(out, F_OK) == 0)
             return;
 
-        /* Strategy 2: sysroot/lib/basename. Handles store-style
-         * interpreter paths such as /.../lib/ld-musl-aarch64.so.1
+        /* Strategy 2: sysroot/lib/basename. Handles store-style interpreter
+         * paths such as /.../lib/ld-musl-aarch64.so.1
          */
         const char *base = strrchr(interp_path, '/');
         base = base ? base + 1 : interp_path;

src/core/elf.h

@@ -4,8 +4,8 @@
  * Copyright 2025 Moritz Angermann, zw3rk pte. ltd.
  * SPDX-License-Identifier: Apache-2.0
  *
- * Parses aarch64-linux ELF64 executables (static and dynamic), extracts
- * PT_LOAD segments, and copies them into guest memory.
+ * Parses aarch64-linux ELF64 executables (static and dynamic), extracts PT_LOAD
+ * segments, and copies them into guest memory.
  */
 
 #pragma once
@@ -107,14 +107,14 @@ typedef struct {
 int elf_load(const char *path, elf_info_t *info);
 
 /* Copy ELF segments into guest memory. Call after elf_load() and guest_init().
- * Also copies program headers into guest memory for AT_PHDR.
- * load_base is added to all virtual addresses (0 for ET_EXEC at link addr,
- * non-zero for ET_DYN loaded at a chosen base).
- * infra_lo and infra_hi delimit the runtime infra reserve (page-table pool,
- * shim text, shim_data, vDSO). Any PT_LOAD or PT_PHDR copy whose destination
- * intersects [infra_lo, infra_hi) is rejected: those writes go through
- * host_base directly and would otherwise bypass the EL1-only page-table
- * protection on shim_data. Pass 0,0 only when the guest_t is not yet built.
+ * Also copies program headers into guest memory for AT_PHDR. load_base is added
+ * to all virtual addresses (0 for ET_EXEC at link addr, non-zero for ET_DYN
+ * loaded at a chosen base). infra_lo and infra_hi delimit the runtime infra
+ * reserve (page-table pool, shim text, shim_data, vDSO). Any PT_LOAD or PT_PHDR
+ * copy whose destination intersects [infra_lo, infra_hi) is rejected: those
+ * writes go through host_base directly and would otherwise bypass the EL1-only
+ * page-table protection on shim_data. Pass 0,0 only when the guest_t is not yet
+ * built.
  * Returns 0 on success, -1 on failure.
  */
 int elf_map_segments(const elf_info_t *info,
@@ -125,8 +125,7 @@ int elf_map_segments(const elf_info_t *info,
                      uint64_t infra_lo,
                      uint64_t infra_hi);
 
-/* Resolve a PT_INTERP path against a sysroot directory.
- * Tries three strategies:
+/* Resolve a PT_INTERP path against a sysroot directory. Tries three strategies:
  *   1. sysroot + interp_path  (standard /lib/ld-musl-*.so.1)
  *   2. sysroot/lib/basename(interp_path)  (store-style paths)
  *   3. interp_path as-is  (no sysroot or fallback)
@@ -138,11 +137,11 @@ void elf_resolve_interp(const char *sysroot,
                         size_t out_sz);
 
 /* Translate ELF program-header flags (PF_R=4, PF_W=2, PF_X=1) into the
- * R=1/W=2/X=4 bitset shared by both MEM_PERM_R/W/X (page-table permissions)
- * and LINUX_PROT_READ/WRITE/EXEC (mmap prot bits).
+ * R=1/W=2/X=4 bitset shared by both MEM_PERM_R/W/X (page-table permissions) and
+ * LINUX_PROT_READ/WRITE/EXEC (mmap prot bits).
  *
- * READ is implicit: every loaded segment gets the R bit even if PF_R is
- * absent, mirroring the kernel's behavior for ELF loading.
+ * READ is implicit: every loaded segment gets the R bit even if PF_R is absent,
+ * mirroring the kernel's behavior for ELF loading.
  */
 static inline int elf_pf_to_prot(int pf)
 {