clamav/libclamav/bytecode.c

/*
 *  Load, and verify ClamAV bytecode.
 *
 *  Copyright (C) 2013-2022 Cisco Systems, Inc. and/or its affiliates. All rights reserved.
 *  Copyright (C) 2009-2013 Sourcefire, Inc.
 *
 *  Authors: Török Edvin
 *
 *  This program is free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License version 2 as
 *  published by the Free Software Foundation.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program; if not, write to the Free Software
 *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
 *  MA 02110-1301, USA.
 */

#if HAVE_CONFIG_H
#include "clamav-config.h"
#endif

#include <string.h>
#include <assert.h>
#include <fcntl.h>

#if HAVE_JSON
#include "json.h"
#endif

#include "dconf.h"
#include "clamav.h"
#include "others.h"
#include "pe.h"
#include "bytecode.h"
#include "bytecode_priv.h"
#include "bytecode_detect.h"
#include "readdb.h"
#include "scanners.h"
#include "bytecode_api.h"
#include "bytecode_api_impl.h"
#include "builtin_bytecodes.h"

#ifndef MAX_TRACKED_BC
#define MAX_TRACKED_BC 64
#endif
#define BC_EVENTS_PER_SIG 2
#define MAX_BC_SIGEVENT_ID MAX_TRACKED_BC *BC_EVENTS_PER_SIG

cli_events_t *g_sigevents = NULL;
unsigned int g_sigid;

/* dummy values */
static const uint32_t nomatch[64] = {
    0xdeadbeef, 0xdeaddead, 0xbeefdead, 0xdeaddead, 0xdeadbeef, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0};
static const uint32_t nooffsets[64] = {
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE,
    CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE, CLI_OFF_NONE};

static const uint16_t nokind;
static const uint32_t nofilesize;
static const struct cli_pe_hook_data nopedata;

static void context_safe(struct cli_bc_ctx *ctx)
{
    /* make sure these are never NULL */
    if (!ctx->hooks.kind)
        ctx->hooks.kind = &nokind;
    if (!ctx->hooks.match_counts)
        ctx->hooks.match_counts = nomatch;
    if (!ctx->hooks.match_offsets)
        ctx->hooks.match_offsets = nooffsets;
    if (!ctx->hooks.filesize)
        ctx->hooks.filesize = &nofilesize;
    if (!ctx->hooks.pedata)
        ctx->hooks.pedata = &nopedata;
}

static int cli_bytecode_context_reset(struct cli_bc_ctx *ctx);
struct cli_bc_ctx *cli_bytecode_context_alloc(void)
{
    struct cli_bc_ctx *ctx = cli_calloc(1, sizeof(*ctx));
    if (!ctx) {
        cli_errmsg("Out of memory allocating cli_bytecode_context_reset\n");
        return NULL;
    }
    ctx->bytecode_timeout = 60000;
    cli_bytecode_context_reset(ctx);
    return ctx;
}

void cli_bytecode_context_destroy(struct cli_bc_ctx *ctx)
{
    cli_bytecode_context_clear(ctx);
    free(ctx);
}

int cli_bytecode_context_getresult_file(struct cli_bc_ctx *ctx, char **tempfilename)
{
    int fd;
    *tempfilename = ctx->tempfile;
    fd            = ctx->outfd;
    ctx->tempfile = NULL;
    ctx->outfd    = 0;
    return fd;
}

/* resets bytecode state, so you can run another bytecode with same ctx */
static int cli_bytecode_context_reset(struct cli_bc_ctx *ctx)
{
    unsigned i;

    free(ctx->opsizes);
    ctx->opsizes = NULL;

    free(ctx->values);
    ctx->values = NULL;

    free(ctx->operands);
    ctx->operands = NULL;

    if (ctx->outfd) {
        cli_ctx *cctx = ctx->ctx;
        if (ctx->outfd)
            close(ctx->outfd);
        if (ctx->tempfile && (!cctx || !cctx->engine->keeptmp)) {
            cli_unlink(ctx->tempfile);
        }
        free(ctx->tempfile);
        ctx->tempfile = NULL;
        ctx->outfd    = 0;
    }
    if (ctx->jsnormdir) {
        char fullname[1025];
        cli_ctx *cctx = ctx->ctx;
        int fd, ret = CL_CLEAN;

        if (!ctx->found) {
            snprintf(fullname, 1024, "%s" PATHSEP "javascript", ctx->jsnormdir);
            fd = open(fullname, O_RDONLY | O_BINARY);
            if (fd >= 0) {
                // This flag ingested by cli_recursion_stack_push().
                cctx->next_layer_attributes |= LAYER_ATTRIBUTES_NORMALIZED;

                ret = cli_scan_desc(fd, cctx, CL_TYPE_HTML, 0, NULL, AC_SCAN_VIR, NULL, NULL);
                if (ret == CL_CLEAN) {
                    if (lseek(fd, 0, SEEK_SET) == -1)
                        cli_dbgmsg("cli_bytecode: call to lseek() has failed\n");
                    else {
                        // This flag ingested by cli_recursion_stack_push().
                        cctx->next_layer_attributes |= LAYER_ATTRIBUTES_NORMALIZED;

                        ret = cli_scan_desc(fd, cctx, CL_TYPE_TEXT_ASCII, 0, NULL, AC_SCAN_VIR, NULL, NULL);
                    }
                }
                close(fd);
            }
        }
        if (!cctx || !cctx->engine->keeptmp) {
            cli_rmdirs(ctx->jsnormdir);
        }
        free(ctx->jsnormdir);
        if (ret != CL_CLEAN)
            ctx->found = 1;
    }
    ctx->numParams = 0;
    ctx->funcid    = 0;
    /* don't touch fmap, file_size, and hooks, sections, ctx, timeout, pdf* */
    ctx->off           = 0;
    ctx->written       = 0;
    ctx->jsnormwritten = 0;
#if USE_MPOOL
    if (ctx->mpool) {
        mpool_destroy(ctx->mpool);
        ctx->mpool = NULL;
    }
#else
    /*TODO: implement for no-mmap case too*/
#endif
    for (i = 0; i < ctx->ninflates; i++)
        cli_bcapi_inflate_done(ctx, i);
    free(ctx->inflates);
    ctx->inflates  = NULL;
    ctx->ninflates = 0;

    for (i = 0; i < ctx->nlzmas; i++)
        cli_bcapi_lzma_done(ctx, i);
    free(ctx->lzmas);
    ctx->lzmas  = NULL;
    ctx->nlzmas = 0;

#if HAVE_BZLIB_H
    for (i = 0; i < ctx->nbzip2s; i++)
        cli_bcapi_bzip2_done(ctx, i);
    free(ctx->bzip2s);
    ctx->bzip2s  = NULL;
    ctx->nbzip2s = 0;
#endif

    for (i = 0; i < ctx->nbuffers; i++)
        cli_bcapi_buffer_pipe_done(ctx, i);
    free(ctx->buffers);
    ctx->buffers  = NULL;
    ctx->nbuffers = 0;

    for (i = 0; i < ctx->nhashsets; i++)
        cli_bcapi_hashset_done(ctx, i);
    free(ctx->hashsets);
    ctx->hashsets  = NULL;
    ctx->nhashsets = 0;

    for (i = 0; i < ctx->njsnorms; i++)
        cli_bcapi_jsnorm_done(ctx, i);
    free(ctx->jsnorms);
    ctx->jsnorms   = NULL;
    ctx->njsnorms  = 0;
    ctx->jsnormdir = NULL;

    for (i = 0; i < ctx->nmaps; i++)
        cli_bcapi_map_done(ctx, i);
    free(ctx->maps);
    ctx->maps  = NULL;
    ctx->nmaps = 0;

    /* Use input_switch() to free the extracted file fmap, if one exists */
    cli_bcapi_input_switch(ctx, 0);

#if HAVE_JSON
    free((json_object **)(ctx->jsonobjs));
    ctx->jsonobjs  = NULL;
    ctx->njsonobjs = 0;
#endif

    ctx->containertype = CL_TYPE_ANY;
    return CL_SUCCESS;
}

int cli_bytecode_context_clear(struct cli_bc_ctx *ctx)
{
    cli_bytecode_context_reset(ctx);
    memset(ctx, 0, sizeof(*ctx));
    return CL_SUCCESS;
}

static unsigned typesize(const struct cli_bc *bc, uint16_t type)
{
    struct cli_bc_type *ty;
    unsigned j;

    type &= 0x7fff;
    if (!type)
        return 0;
    if (type <= 8)
        return 1;
    if (type <= 16)
        return 2;
    if (type <= 32)
        return 4;
    if (type <= 64)
        return 8;
    ty = &bc->types[type - 65];
    if (ty->size)
        return ty->size;
    switch (ty->kind) {
        case 2:
        case 3:
            for (j = 0; j < ty->numElements; j++)
                ty->size += typesize(bc, ty->containedTypes[j]);
            break;
        case 4:
            ty->size = ty->numElements * typesize(bc, ty->containedTypes[0]);
            break;
        default:
            break;
    }
    if (!ty->size && ty->kind != DFunctionType) {
        cli_warnmsg("type %d size is 0\n", type - 65);
    }
    return ty->size;
}

static unsigned typealign(const struct cli_bc *bc, uint16_t type)
{
    type &= 0x7fff;
    if (type <= 64) {
        unsigned size = typesize(bc, type);
        return size ? size : 1;
    }
    return bc->types[type - 65].align;
}

int cli_bytecode_context_setfuncid(struct cli_bc_ctx *ctx, const struct cli_bc *bc, unsigned funcid)
{
    unsigned i, s = 0;
    const struct cli_bc_func *func;
    if (funcid >= bc->num_func) {
        cli_errmsg("bytecode: function ID doesn't exist: %u\n", funcid);
        return CL_EARG;
    }
    func = ctx->func = &bc->funcs[funcid];
    ctx->bc          = bc;
    ctx->numParams   = func->numArgs;
    ctx->funcid      = funcid;
    if (func->numArgs) {
        ctx->operands = cli_malloc(sizeof(*ctx->operands) * func->numArgs);
        if (!ctx->operands) {
            cli_errmsg("bytecode: error allocating memory for parameters\n");
            return CL_EMEM;
        }
        ctx->opsizes = cli_malloc(sizeof(*ctx->opsizes) * func->numArgs);
        if (!ctx->opsizes) {
            cli_errmsg("bytecode: error allocating memory for opsizes\n");
            return CL_EMEM;
        }
        for (i = 0; i < func->numArgs; i++) {
            unsigned al          = typealign(bc, func->types[i]);
            s                    = (s + al - 1) & ~(al - 1);
            ctx->operands[i]     = s;
            s += ctx->opsizes[i] = typesize(bc, func->types[i]);
        }
    }
    s += 8; /* return value */
    ctx->bytes  = s;
    ctx->values = cli_malloc(s);
    if (!ctx->values) {
        cli_errmsg("bytecode: error allocating memory for parameters\n");
        return CL_EMEM;
    }
    return CL_SUCCESS;
}

static inline int type_isint(uint16_t type)
{
    return type > 0 && type <= 64;
}

int cli_bytecode_context_setparam_int(struct cli_bc_ctx *ctx, unsigned i, uint64_t c)
{
    if (i >= ctx->numParams) {
        cli_errmsg("bytecode: param index out of bounds: %u\n", i);
        return CL_EARG;
    }
    if (!type_isint(ctx->func->types[i])) {
        cli_errmsg("bytecode: parameter type mismatch\n");
        return CL_EARG;
    }
    switch (ctx->opsizes[i]) {
        case 1:
            ctx->values[ctx->operands[i]] = c;
            break;
        case 2:
            *(uint16_t *)&ctx->values[ctx->operands[i]] = c;
            break;
        case 4:
            *(uint32_t *)&ctx->values[ctx->operands[i]] = c;
            break;
        case 8:
            *(uint64_t *)&ctx->values[ctx->operands[i]] = c;
            break;
    }
    return CL_SUCCESS;
}

int cli_bytecode_context_setparam_ptr(struct cli_bc_ctx *ctx, unsigned i, void *data, unsigned datalen)
{
    UNUSEDPARAM(ctx);
    UNUSEDPARAM(i);
    UNUSEDPARAM(data);
    UNUSEDPARAM(datalen);
    cli_errmsg("Pointer parameters are not implemented yet!\n");
    return CL_EARG;
}

static inline uint64_t readNumber(const unsigned char *p, unsigned *off, unsigned len, char *ok)
{
    uint64_t n = 0;
    unsigned i, newoff, lim, p0 = p[*off], shift = 0;

    lim = p0 - 0x60;
    if (lim > 0x10) {
        cli_errmsg("Invalid number type: %c\n", p0);
        *ok = 0;
        return 0;
    }
    newoff = *off + lim + 1;
    if (newoff > len) {
        cli_errmsg("End of line encountered while reading number\n");
        *ok = 0;
        return 0;
    }

    if (p0 == 0x60) {
        *off = newoff;
        return 0;
    }

    for (i = *off + 1; i < newoff; i++) {
        uint64_t v = p[i];
        if (UNLIKELY((v & 0xf0) != 0x60)) {
            cli_errmsg("Invalid number part: %c\n", (char)v);
            *ok = 0;
            return 0;
        }
        v &= 0xf;
        v <<= shift;
        n |= v;
        shift += 4;
    }
    *off = newoff;
    return n;
}

static inline funcid_t readFuncID(struct cli_bc *bc, unsigned char *p,
                                  unsigned *off, unsigned len, char *ok)
{
    funcid_t id = readNumber(p, off, len, ok) - 1;
    if (*ok && id >= bc->num_func) {
        cli_errmsg("Called function out of range: %u >= %u\n", id, bc->num_func);
        *ok = 0;
        return ~0;
    }
    return id;
}

static inline funcid_t readAPIFuncID(struct cli_bc *bc, unsigned char *p,
                                     unsigned *off, unsigned len, char *ok)
{
    funcid_t id = readNumber(p, off, len, ok) - 1;
    if (*ok && !cli_bitset_test(bc->uses_apis, id)) {
        cli_errmsg("Called undeclared API function: %u\n", id);
        *ok = 0;
        return ~0;
    }
    return id;
}

static inline unsigned readFixedNumber(const unsigned char *p, unsigned *off,
                                       unsigned len, char *ok, unsigned width)
{
    unsigned i, n = 0, shift = 0;
    unsigned newoff = *off + width;
    if (newoff > len) {
        cli_errmsg("Newline encountered while reading number\n");
        *ok = 0;
        return 0;
    }
    for (i = *off; i < newoff; i++) {
        unsigned v = p[i];
        if (UNLIKELY((v & 0xf0) != 0x60)) {
            cli_errmsg("Invalid number part: %c\n", v);
            *ok = 0;
            return 0;
        }
        v &= 0xf;
        v <<= shift;
        n |= v;
        shift += 4;
    }
    *off = newoff;
    return n;
}

static inline operand_t readOperand(struct cli_bc_func *func, unsigned char *p,
                                    unsigned *off, unsigned len, char *ok)
{
    uint64_t v;
    if ((p[*off] & 0xf0) == 0x40 || p[*off] == 0x50) {
        uint64_t *dest;
        uint16_t ty;
        p[*off] |= 0x20;
        /* TODO: unique constants */
        func->constants = cli_realloc2(func->constants, (func->numConstants + 1) * sizeof(*func->constants));
        if (!func->constants) {
            *ok = 0;
            return MAX_OP;
        }
        v    = readNumber(p, off, len, ok);
        dest = &func->constants[func->numConstants];
        /* Write the constant to the correct place according to its type.
         * This is needed on big-endian machines, because constants are always
         * read as u64, but accessed as one of these types: u8, u16, u32, u64 */
        *dest = 0;
        ty    = 8 * readFixedNumber(p, off, len, ok, 1);
        if (!ty) {
            /* This is a global variable */
            return 0x80000000 | v;
        }
        if (ty <= 8)
            *(uint8_t *)dest = v;
        else if (ty <= 16)
            *(uint16_t *)dest = v;
        else if (ty <= 32)
            *(uint32_t *)dest = v;
        else
            *dest = v;
        return func->numValues + func->numConstants++;
    }
    v = readNumber(p, off, len, ok);
    if (!*ok)
        return MAX_OP;
    if (v >= func->numValues) {
        cli_errmsg("Operand index exceeds bounds: %u >= %u!\n", (unsigned)v, (unsigned)func->numValues);
        *ok = 0;
        return MAX_OP;
    }
    return v;
}

static inline char *readData(const unsigned char *p, unsigned *off, unsigned len, char *ok, unsigned *datalen)
{
    unsigned char *dat, *q;
    unsigned l, newoff, i;
    if (p[*off] != '|') {
        cli_errmsg("Data start marker missing: %c\n", p[*off]);
        *ok = 0;
        return NULL;
    }
    (*off)++;
    l = readNumber(p, off, len, ok);
    if (!l || !ok) {
        *datalen = l;
        return NULL;
    }
    newoff = *off + 2 * l;
    if (newoff > len) {
        cli_errmsg("Line ended while reading data\n");
        *ok = 0;
        return 0;
    }
    dat = cli_malloc(l);
    if (!dat) {
        cli_errmsg("Cannot allocate memory for data\n");
        *ok = 0;
        return NULL;
    }
    q = dat;
    for (i = *off; i < newoff; i += 2) {
        const unsigned char v0 = p[i];
        const unsigned char v1 = p[i + 1];
        if (UNLIKELY((v0 & 0xf0) != 0x60 || (v1 & 0xf0) != 0x60)) {
            cli_errmsg("Invalid data part: %c%c\n", v0, v1);
            *ok = 0;
            free(dat);
            return 0;
        }
        *q++ = (v0 & 0xf) | ((v1 & 0xf) << 4);
    }
    *off     = newoff;
    *datalen = l;
    return (char *)dat;
}

static inline char *readString(const unsigned char *p, unsigned *off, unsigned len, char *ok)
{
    unsigned stringlen = 0;
    char *str          = readData(p, off, len, ok, &stringlen);
    if (*ok && stringlen && str[stringlen - 1] != '\0') {
        str[stringlen - 1] = '\0';
        cli_errmsg("bytecode: string missing \\0 terminator: %s\n", str);
        free(str);
        *ok = 0;
        return NULL;
    }
    return str;
}

static int parseHeader(struct cli_bc *bc, unsigned char *buffer, unsigned *linelength)
{
    uint64_t magic1;
    unsigned magic2;
    char ok = 1;
    unsigned offset, len, flevel;
    char *pos;

    if (strncmp((const char *)buffer, BC_HEADER, sizeof(BC_HEADER) - 1)) {
        cli_errmsg("Missing file magic in bytecode");
        return CL_EMALFDB;
    }
    offset                   = sizeof(BC_HEADER) - 1;
    len                      = strlen((const char *)buffer);
    bc->metadata.formatlevel = readNumber(buffer, &offset, len, &ok);
    if (!ok) {
        cli_errmsg("Unable to parse (format) functionality level in bytecode header\n");
        return CL_EMALFDB;
    }
    /* we support 2 bytecode formats */
    if (bc->metadata.formatlevel != BC_FORMAT_096 &&
        bc->metadata.formatlevel != BC_FORMAT_LEVEL) {
        cli_dbgmsg("Skipping bytecode with (format) functionality level: %u (current %u)\n",
                   bc->metadata.formatlevel, BC_FORMAT_LEVEL);
        return CL_BREAK;
    }
    /* Optimistic parsing, check for error only at the end.*/
    bc->metadata.timestamp     = readNumber(buffer, &offset, len, &ok);
    bc->metadata.sigmaker      = readString(buffer, &offset, len, &ok);
    bc->metadata.targetExclude = readNumber(buffer, &offset, len, &ok);
    bc->kind                   = readNumber(buffer, &offset, len, &ok);
    bc->metadata.minfunc       = readNumber(buffer, &offset, len, &ok);
    bc->metadata.maxfunc       = readNumber(buffer, &offset, len, &ok);
    flevel                     = cl_retflevel();
    /* in 0.96 these 2 fields are unused / zero, in post 0.96 these mean
     * min/max flevel.
     * So 0 for min/max means no min/max
     * Note that post 0.96 bytecode/bytecode lsig needs format 7, because
     * 0.96 doesn't check lsig functionality level.
     */
    if ((bc->metadata.minfunc && bc->metadata.minfunc > flevel) ||
        (bc->metadata.maxfunc && bc->metadata.maxfunc < flevel)) {
        cli_dbgmsg("Skipping bytecode with (engine) functionality level %u-%u (current %u)\n",
                   bc->metadata.minfunc, bc->metadata.maxfunc, flevel);
        return CL_BREAK;
    }
    bc->metadata.maxresource = readNumber(buffer, &offset, len, &ok);
    bc->metadata.compiler    = readString(buffer, &offset, len, &ok);
    bc->num_types            = readNumber(buffer, &offset, len, &ok);
    bc->num_func             = readNumber(buffer, &offset, len, &ok);
    bc->state                = bc_loaded;
    bc->uses_apis            = NULL;
    bc->dbgnodes             = NULL;
    bc->dbgnode_cnt          = 0;
    if (!ok) {
        cli_errmsg("Invalid bytecode header at %u\n", offset);
        return CL_EMALFDB;
    }
    magic1 = readNumber(buffer, &offset, len, &ok);
    magic2 = readFixedNumber(buffer, &offset, len, &ok, 2);
    if (!ok || magic1 != 0x53e5493e9f3d1c30ull || magic2 != 42) {
        unsigned long m0 = magic1 >> 32;
        unsigned long m1 = magic1;
        cli_errmsg("Magic numbers don't match: %lx%lx, %u\n", m0, m1, magic2);
        return CL_EMALFDB;
    }
    if (buffer[offset] != ':') {
        cli_errmsg("Expected : but found: %c\n", buffer[offset]);
        return CL_EMALFDB;
    }
    offset++;
    *linelength = strtol((const char *)buffer + offset, &pos, 10);
    if (*pos != '\0') {
        cli_errmsg("Invalid number: %s\n", buffer + offset);
        return CL_EMALFDB;
    }

    bc->funcs = cli_calloc(bc->num_func, sizeof(*bc->funcs));
    if (!bc->funcs) {
        cli_errmsg("Out of memory allocating %u functions\n", bc->num_func);
        return CL_EMEM;
    }
    bc->types = cli_calloc(bc->num_types, sizeof(*bc->types));
    if (!bc->types) {
        cli_errmsg("Out of memory allocating %u types\n", bc->num_types);
        return CL_EMEM;
    }
    return CL_SUCCESS;
}

static int parseLSig(struct cli_bc *bc, char *buffer)
{
    const char *prefix;
    char *vnames, *vend = strchr(buffer, ';');
    if (vend) {
        bc->lsig = cli_strdup(buffer);
        *vend++  = '\0';
        prefix   = buffer;
        vnames   = strchr(vend, '{');
    } else {
        /* Not a logical signature, but we still have a virusname */
        bc->hook_name = cli_strdup(buffer);
        bc->lsig      = NULL;
    }

    return CL_SUCCESS;
}

static uint16_t readTypeID(struct cli_bc *bc, unsigned char *buffer,
                           unsigned *offset, unsigned len, char *ok)
{
    uint64_t t = readNumber(buffer, offset, len, ok);
    if (!ok)
        return ~0;
    if (t >= bc->num_types + bc->start_tid) {
        cli_errmsg("Invalid type id: %llu\n", (unsigned long long)t);
        *ok = 0;
        return ~0;
    }
    return t;
}

static void parseType(struct cli_bc *bc, struct cli_bc_type *ty,
                      unsigned char *buffer, unsigned *off, unsigned len,
                      char *ok)
{
    unsigned j;

    ty->numElements = readNumber(buffer, off, len, ok);
    if (!*ok) {
        cli_errmsg("Error parsing type\n");
        *ok = 0;
        return;
    }
    ty->containedTypes = cli_malloc(sizeof(*ty->containedTypes) * ty->numElements);
    if (!ty->containedTypes) {
        cli_errmsg("Out of memory allocating %u types\n", ty->numElements);
        *ok = 0;
        return;
    }
    for (j = 0; j < ty->numElements; j++) {
        ty->containedTypes[j] = readTypeID(bc, buffer, off, len, ok);
    }
}

static uint16_t containedTy[] = {8, 16, 32, 64};

#define NUM_STATIC_TYPES 4
static void add_static_types(struct cli_bc *bc)
{
    unsigned i;
    for (i = 0; i < NUM_STATIC_TYPES; i++) {
        bc->types[i].kind           = DPointerType;
        bc->types[i].numElements    = 1;
        bc->types[i].containedTypes = &containedTy[i];
        bc->types[i].size = bc->types[i].align = 8;
    }
}

static int parseTypes(struct cli_bc *bc, unsigned char *buffer)
{
    unsigned i, offset = 1, len = strlen((const char *)buffer);
    char ok = 1;

    if (buffer[0] != 'T') {
        cli_errmsg("Invalid function types header: %c\n", buffer[0]);
        return CL_EMALFDB;
    }
    bc->start_tid = readFixedNumber(buffer, &offset, len, &ok, 2);
    if (bc->start_tid != BC_START_TID) {
        cli_warnmsg("Type start id mismatch: %u != %u\n", bc->start_tid,
                    BC_START_TID);
        return CL_BREAK;
    }
    add_static_types(bc);
    for (i = (BC_START_TID - 65); i < bc->num_types - 1; i++) {
        struct cli_bc_type *ty = &bc->types[i];
        uint8_t t              = readFixedNumber(buffer, &offset, len, &ok, 1);
        if (!ok) {
            cli_errmsg("Error reading type kind\n");
            return CL_EMALFDB;
        }
        switch (t) {
            case 1:
                ty->kind = DFunctionType;
                ty->size = ty->align = sizeof(void *);
                parseType(bc, ty, buffer, &offset, len, &ok);
                if (!ok) {
                    cli_errmsg("Error parsing type %u\n", i);
                    return CL_EMALFDB;
                }
                if (!ty->numElements) {
                    cli_errmsg("Function with no return type? %u\n", i);
                    return CL_EMALFDB;
                }
                break;
            case 2:
            case 3:
                ty->kind = (t == 2) ? DPackedStructType : DStructType;
                ty->size = ty->align = 0; /* TODO:calculate size/align of structs */
                ty->align            = 8;
                parseType(bc, ty, buffer, &offset, len, &ok);
                if (!ok) {
                    cli_errmsg("Error parsing type %u\n", i);
                    return CL_EMALFDB;
                }
                break;
            case 4:
                ty->kind = DArrayType;
                /* number of elements of array, not subtypes! */
                ty->numElements = readNumber(buffer, &offset, len, &ok);
                if (!ok) {
                    cli_errmsg("Error parsing type %u\n", i);
                    return CL_EMALFDB;
                }
                /* fall-through */
            case 5:
                if (t == 5) {
                    ty->kind        = DPointerType;
                    ty->numElements = 1;
                }
                ty->containedTypes = cli_malloc(sizeof(*ty->containedTypes));
                if (!ty->containedTypes) {
                    cli_errmsg("Out of memory allocating containedType\n");
                    return CL_EMALFDB;
                }
                ty->containedTypes[0] = readTypeID(bc, buffer, &offset, len, &ok);
                if (!ok) {
                    cli_errmsg("Error parsing type %u\n", i);
                    return CL_EMALFDB;
                }
                if (t == 5) {
                    /* for interpreter, pointers 64-bit there */
                    ty->size = ty->align = 8;
                } else {
                    ty->size  = ty->numElements * typesize(bc, ty->containedTypes[0]);
                    ty->align = typealign(bc, ty->containedTypes[0]);
                }
                break;
            default:
                cli_errmsg("Invalid type kind: %u\n", t);
                return CL_EMALFDB;
        }
    }
    for (i = (BC_START_TID - 65); i < bc->num_types - 1; i++) {
        struct cli_bc_type *ty = &bc->types[i];
        if (ty->kind == DArrayType) {
            ty->size  = ty->numElements * typesize(bc, ty->containedTypes[0]);
            ty->align = typealign(bc, ty->containedTypes[0]);
        }
    }
    return CL_SUCCESS;
}

/* checks whether the type described by tid is the same as the one described by
 * apitid. */
static int types_equal(const struct cli_bc *bc, uint16_t *apity2ty, uint16_t tid, uint16_t apitid)
{
    unsigned i;
    const struct cli_bc_type *ty    = &bc->types[tid - 65];
    const struct cli_bc_type *apity = &cli_apicall_types[apitid];
    /* If we've already verified type equality, return.
     * Since we need to check equality of recursive types, we assume types are
     * equal while checking equality of contained types, unless proven
     * otherwise. */
    if (apity2ty[apitid] == tid + 1)
        return 1;
    apity2ty[apitid] = tid + 1;

    if (ty->kind != apity->kind) {
        cli_dbgmsg("bytecode: type kind mismatch: %u != %u\n", ty->kind, apity->kind);
        return 0;
    }
    if (ty->numElements != apity->numElements) {
        cli_dbgmsg("bytecode: type numElements mismatch: %u != %u\n", ty->numElements, apity->numElements);
        return 0;
    }
    for (i = 0; i < ty->numElements; i++) {
        if (apity->containedTypes[i] < BC_START_TID) {
            if (ty->containedTypes[i] != apity->containedTypes[i]) {
                cli_dbgmsg("bytecode: contained type mismatch: %u != %u\n",
                           ty->containedTypes[i], apity->containedTypes[i]);
                return 0;
            }
        } else if (!types_equal(bc, apity2ty, ty->containedTypes[i], apity->containedTypes[i] - BC_START_TID))
            return 0;
        if (ty->kind == DArrayType)
            break; /* validated the contained type already */
    }
    return 1;
}

static int parseApis(struct cli_bc *bc, unsigned char *buffer)
{
    unsigned i, offset = 1, len = strlen((const char *)buffer), maxapi, calls;
    char ok = 1;
    uint16_t *apity2ty; /*map of api type to current bytecode type ID */

    if (buffer[0] != 'E') {
        cli_errmsg("bytecode: Invalid api header: %c\n", buffer[0]);
        return CL_EMALFDB;
    }

    maxapi = readNumber(buffer, &offset, len, &ok);
    if (!ok)
        return CL_EMALFDB;
    if (maxapi > cli_apicall_maxapi) {
        cli_dbgmsg("bytecode using API %u, but highest API known to libclamav is %u, skipping\n", maxapi, cli_apicall_maxapi);
        return CL_BREAK;
    }
    calls = readNumber(buffer, &offset, len, &ok);
    if (!ok)
        return CL_EMALFDB;
    if (calls > maxapi) {
        cli_errmsg("bytecode: attempting to describe more APIs than max: %u > %u\n", calls, maxapi);
        return CL_EMALFDB;
    }
    bc->uses_apis = cli_bitset_init();
    if (!bc->uses_apis) {
        cli_errmsg("Out of memory allocating apis bitset\n");
        return CL_EMEM;
    }
    apity2ty = cli_calloc(cli_apicall_maxtypes, sizeof(*cli_apicall_types));
    if (!apity2ty) {
        cli_errmsg("Out of memory allocating apity2ty\n");
        return CL_EMEM;
    }
    for (i = 0; i < calls; i++) {
        unsigned id  = readNumber(buffer, &offset, len, &ok);
        uint16_t tid = readTypeID(bc, buffer, &offset, len, &ok);
        char *name   = readString(buffer, &offset, len, &ok);

        /* validate APIcall prototype */
        if (id > maxapi) {
            cli_errmsg("bytecode: API id %u out of range, max %u\n", id, maxapi);
            ok = 0;
        }
        /* API ids start from 1 */
        id--;
        if (ok && name && strcmp(cli_apicalls[id].name, name)) {
            cli_errmsg("bytecode: API %u name mismatch: %s expected %s\n", id, name, cli_apicalls[id].name);
            ok = 0;
        }
        if (ok && !types_equal(bc, apity2ty, tid, cli_apicalls[id].type)) {
            cli_errmsg("bytecode: API %u prototype doesn't match\n", id);
            ok = 0;
        }
        /* don't need the name anymore */
        free(name);
        if (!ok) {
            free(apity2ty); /* free temporary map */
            return CL_EMALFDB;
        }

        /* APIcall is valid */
        cli_bitset_set(bc->uses_apis, id);
    }
    free(apity2ty); /* free temporary map */
    cli_dbgmsg("bytecode: Parsed %u APIcalls, maxapi %u\n", calls, maxapi);
    return CL_SUCCESS;
}

static uint16_t type_components(struct cli_bc *bc, uint16_t id, char *ok)
{
    unsigned i, sum = 0;
    const struct cli_bc_type *ty;
    if (id <= 64)
        return 1;
    ty = &bc->types[id - 65];
    /* TODO: protect against recursive types */
    switch (ty->kind) {
        case DFunctionType:
            cli_errmsg("bytecode: function type not accepted for constant: %u\n", id);
            /* don't accept functions as constant initializers */
            *ok = 0;
            return 0;
        case DPointerType:
            return 2;
        case DStructType:
        case DPackedStructType:
            for (i = 0; i < ty->numElements; i++) {
                sum += type_components(bc, ty->containedTypes[i], ok);
            }
            return sum;
        case DArrayType:
            return type_components(bc, ty->containedTypes[0], ok) * ty->numElements;
        default:
            *ok = 0;
            return 0;
    }
}

static void readConstant(struct cli_bc *bc, unsigned i, unsigned comp,
                         unsigned char *buffer, unsigned *offset,
                         unsigned len, char *ok)
{
    unsigned j = 0;
    if (*ok && buffer[*offset] == 0x40 &&
        buffer[*offset + 1] == 0x60) {
        /* zero initializer */
        memset(bc->globals[i], 0, sizeof(*bc->globals[0]) * comp);
        (*offset) += 2;
        return;
    }
    while (*ok && buffer[*offset] != 0x60) {
        if (j >= comp) {
            cli_errmsg("bytecode: constant has too many subcomponents, expected %u\n", comp);
            *ok = 0;
            return;
        }
        buffer[*offset] |= 0x20;
        bc->globals[i][j++] = readNumber(buffer, offset, len, ok);
    }
    if (*ok && j != comp) {
        cli_errmsg("bytecode: constant has too few subcomponents: %u < %u\n", j, comp);
        *ok = 0;
    }
    (*offset)++;
}

/* parse constant globals with constant initializers */
static int parseGlobals(struct cli_bc *bc, unsigned char *buffer)
{
    unsigned i, offset = 1, len = strlen((const char *)buffer), numglobals;
    unsigned maxglobal;
    char ok = 1;

    if (buffer[0] != 'G') {
        cli_errmsg("bytecode: Invalid globals header: %c\n", buffer[0]);
        return CL_EMALFDB;
    }
    maxglobal = readNumber(buffer, &offset, len, &ok);
    if (maxglobal > cli_apicall_maxglobal) {
        cli_dbgmsg("bytecode using global %u, but highest global known to libclamav is %u, skipping\n", maxglobal, cli_apicall_maxglobal);
        return CL_BREAK;
    }
    numglobals  = readNumber(buffer, &offset, len, &ok);
    bc->globals = cli_calloc(numglobals, sizeof(*bc->globals));
    if (!bc->globals) {
        cli_errmsg("bytecode: OOM allocating memory for %u globals\n", numglobals);
        return CL_EMEM;
    }
    bc->globaltys = cli_calloc(numglobals, sizeof(*bc->globaltys));
    if (!bc->globaltys) {
        cli_errmsg("bytecode: OOM allocating memory for %u global types\n", numglobals);
        return CL_EMEM;
    }
    bc->num_globals = numglobals;
    if (!ok)
        return CL_EMALFDB;
    for (i = 0; i < numglobals; i++) {
        unsigned comp;
        bc->globaltys[i] = readTypeID(bc, buffer, &offset, len, &ok);
        comp             = type_components(bc, bc->globaltys[i], &ok);
        if (!ok)
            return CL_EMALFDB;
        bc->globals[i] = cli_malloc(sizeof(*bc->globals[0]) * comp);
        if (!bc->globals[i])
            return CL_EMEM;
        readConstant(bc, i, comp, buffer, &offset, len, &ok);
    }
    if (!ok)
        return CL_EMALFDB;
    if (offset != len) {
        cli_errmsg("Trailing garbage in globals: %d extra bytes\n",
                   len - offset);
        return CL_EMALFDB;
    }
    return CL_SUCCESS;
}

static int parseMD(struct cli_bc *bc, unsigned char *buffer)
{
    unsigned offset = 1, len = strlen((const char *)buffer);
    unsigned numMD, i, b;
    char ok = 1;
    if (buffer[0] != 'D')
        return CL_EMALFDB;
    numMD = readNumber(buffer, &offset, len, &ok);
    if (!ok) {
        cli_errmsg("Unable to parse number of MD nodes\n");
        return CL_EMALFDB;
    }
    b = bc->dbgnode_cnt;
    bc->dbgnode_cnt += numMD;
    bc->dbgnodes = cli_realloc(bc->dbgnodes, bc->dbgnode_cnt * sizeof(*bc->dbgnodes));
    if (!bc->dbgnodes)
        return CL_EMEM;
    for (i = 0; i < numMD; i++) {
        unsigned j;
        struct cli_bc_dbgnode_element *elts;
        unsigned el = readNumber(buffer, &offset, len, &ok);
        if (!ok) {
            cli_errmsg("Unable to parse number of elements\n");
            return CL_EMALFDB;
        }
        bc->dbgnodes[b + i].numelements = el;
        bc->dbgnodes[b + i].elements = elts = cli_calloc(el, sizeof(*elts));
        if (!elts)
            return CL_EMEM;
        for (j = 0; j < el; j++) {
            if (buffer[offset] == '|') {
                elts[j].string = readData(buffer, &offset, len, &ok, &elts[j].len);
                if (!ok)
                    return CL_EMALFDB;
            } else {
                elts[j].len = readNumber(buffer, &offset, len, &ok);
                if (!ok)
                    return CL_EMALFDB;
                if (elts[j].len) {
                    elts[j].constant = readNumber(buffer, &offset, len, &ok);
                } else
                    elts[j].nodeid = readNumber(buffer, &offset, len, &ok);
                if (!ok)
                    return CL_EMALFDB;
            }
        }
    }
    cli_dbgmsg("bytecode: Parsed %u nodes total\n", bc->dbgnode_cnt);
    return CL_SUCCESS;
}

static int parseFunctionHeader(struct cli_bc *bc, unsigned fn, unsigned char *buffer)
{
    char ok = 1;
    unsigned offset, len, all_locals = 0, i;
    struct cli_bc_func *func;

    if (fn >= bc->num_func) {
        cli_errmsg("Found more functions than declared: %u >= %u\n", fn,
                   bc->num_func);
        return CL_EMALFDB;
    }
    func = &bc->funcs[fn];
    len  = strlen((const char *)buffer);

    if (buffer[0] != 'A') {
        cli_errmsg("Invalid function arguments header: %c\n", buffer[0]);
        return CL_EMALFDB;
    }
    offset           = 1;
    func->numArgs    = readFixedNumber(buffer, &offset, len, &ok, 1);
    func->returnType = readTypeID(bc, buffer, &offset, len, &ok);
    if (buffer[offset] != 'L') {
        cli_errmsg("Invalid function locals header: %c\n", buffer[offset]);
        return CL_EMALFDB;
    }
    offset++;
    func->numLocals = readNumber(buffer, &offset, len, &ok);
    if (!ok) {
        cli_errmsg("Invalid number of arguments/locals\n");
        return CL_EMALFDB;
    }
    all_locals = func->numArgs + func->numLocals;
    if (!all_locals) {
        func->types = NULL;
    } else {
        func->types = cli_calloc(all_locals, sizeof(*func->types));
        if (!func->types) {
            cli_errmsg("Out of memory allocating function arguments\n");
            return CL_EMEM;
        }
    }
    for (i = 0; i < all_locals; i++) {
        func->types[i] = readNumber(buffer, &offset, len, &ok);
        if (readFixedNumber(buffer, &offset, len, &ok, 1))
            func->types[i] |= 0x8000;
    }
    if (!ok) {
        cli_errmsg("Invalid local types\n");
        return CL_EMALFDB;
    }
    if (buffer[offset] != 'F') {
        cli_errmsg("Invalid function body header: %c\n", buffer[offset]);
        return CL_EMALFDB;
    }
    offset++;
    func->numInsts = readNumber(buffer, &offset, len, &ok);
    if (!ok) {
        cli_errmsg("Invalid instructions count\n");
        return CL_EMALFDB;
    }
    func->numValues    = func->numArgs + func->numLocals;
    func->insn_idx     = 0;
    func->numConstants = 0;
    func->allinsts     = cli_calloc(func->numInsts, sizeof(*func->allinsts));
    if (!func->allinsts) {
        cli_errmsg("Out of memory allocating instructions\n");
        return CL_EMEM;
    }
    func->numBB = readNumber(buffer, &offset, len, &ok);
    if (!ok) {
        cli_errmsg("Invalid basic block count\n");
        return CL_EMALFDB;
    }
    func->BB = cli_calloc(func->numBB, sizeof(*func->BB));
    if (!func->BB) {
        cli_errmsg("Out of memory allocating basic blocks\n");
        return CL_EMEM;
    }
    return CL_SUCCESS;
}

static bbid_t readBBID(struct cli_bc_func *func, const unsigned char *buffer, unsigned *off, unsigned len, char *ok)
{
    unsigned id = readNumber(buffer, off, len, ok);
    if (!id || id >= func->numBB) {
        cli_errmsg("Basic block ID out of range: %u\n", id);
        *ok = 0;
    }
    if (!*ok)
        return ~0;
    return id;
}

/*
static uint16_t get_type(struct cli_bc_func *func, operand_t op)
{
    if (op >= func->numValues)
        return 64;
    return func->types[op];
}*/
static int16_t get_optype(const struct cli_bc_func *bcfunc, operand_t op)
{
    if (op >= bcfunc->numArgs + bcfunc->numLocals)
        return 0;
    return bcfunc->types[op] & 0x7fff;
}

static int parseBB(struct cli_bc *bc, unsigned func, unsigned bb, unsigned char *buffer)
{
    char ok = 1;
    unsigned offset, len, i, last = 0;
    struct cli_bc_bb *BB;
    struct cli_bc_func *bcfunc = &bc->funcs[func];
    struct cli_bc_inst inst;

    if (bb >= bcfunc->numBB) {
        cli_errmsg("Found too many basic blocks\n");
        return CL_EMALFDB;
    }

    BB  = &bcfunc->BB[bb];
    len = strlen((const char *)buffer);
    if (buffer[0] != 'B') {
        cli_errmsg("Invalid basic block header: %c\n", buffer[0]);
        return CL_EMALFDB;
    }
    offset       = 1;
    BB->numInsts = 0;
    BB->insts    = &bcfunc->allinsts[bcfunc->insn_idx];
    while (!last) {
        unsigned numOp;
        if (buffer[offset] == 'T') {
            last = 1;
            offset++;
            /* terminators are void */
            inst.type = 0;
            inst.dest = 0;
        } else {
            inst.type = readNumber(buffer, &offset, len, &ok);
            inst.dest = readNumber(buffer, &offset, len, &ok);
        }
        inst.opcode = readFixedNumber(buffer, &offset, len, &ok, 2);
        if (!ok) {
            cli_errmsg("Invalid type or operand\n");
            return CL_EMALFDB;
        }
        if (inst.opcode >= OP_BC_INVALID) {
            cli_errmsg("Invalid opcode: %u\n", inst.opcode);
            return CL_EMALFDB;
        }

        switch (inst.opcode) {
            case OP_BC_JMP:
                inst.u.jump = readBBID(bcfunc, buffer, &offset, len, &ok);
                break;
            case OP_BC_RET:
                inst.type      = readNumber(buffer, &offset, len, &ok);
                inst.u.unaryop = readOperand(bcfunc, buffer, &offset, len, &ok);
                break;
            case OP_BC_BRANCH:
                inst.u.branch.condition = readOperand(bcfunc, buffer, &offset, len, &ok);
                inst.u.branch.br_true   = readBBID(bcfunc, buffer, &offset, len, &ok);
                inst.u.branch.br_false  = readBBID(bcfunc, buffer, &offset, len, &ok);
                break;
            case OP_BC_CALL_API: /* fall-through */
            case OP_BC_CALL_DIRECT:
                numOp = readFixedNumber(buffer, &offset, len, &ok, 1);
                if (ok) {
                    inst.u.ops.numOps  = numOp;
                    inst.u.ops.opsizes = NULL;
                    if (!numOp) {
                        inst.u.ops.ops = NULL;
                    } else {
                        inst.u.ops.ops = cli_calloc(numOp, sizeof(*inst.u.ops.ops));
                        if (!inst.u.ops.ops) {
                            cli_errmsg("Out of memory allocating operands\n");
                            return CL_EMEM;
                        }
                    }
                    if (inst.opcode == OP_BC_CALL_DIRECT)
                        inst.u.ops.funcid = readFuncID(bc, buffer, &offset, len, &ok);
                    else
                        inst.u.ops.funcid = readAPIFuncID(bc, buffer, &offset, len, &ok);
                    for (i = 0; i < numOp; i++) {
                        inst.u.ops.ops[i] = readOperand(bcfunc, buffer, &offset, len, &ok);
                    }
                }
                break;
            case OP_BC_ZEXT:
            case OP_BC_SEXT:
            case OP_BC_TRUNC:
                inst.u.cast.source = readOperand(bcfunc, buffer, &offset, len, &ok);
                inst.u.cast.mask   = bcfunc->types[inst.u.cast.source];
                if (inst.u.cast.mask == 1)
                    inst.u.cast.size = 0;
                else if (inst.u.cast.mask <= 8)
                    inst.u.cast.size = 1;
                else if (inst.u.cast.mask <= 16)
                    inst.u.cast.size = 2;
                else if (inst.u.cast.mask <= 32)
                    inst.u.cast.size = 3;
                else if (inst.u.cast.mask <= 64)
                    inst.u.cast.size = 4;
                /* calculate mask */
                if (inst.opcode != OP_BC_SEXT)
                    inst.u.cast.mask = inst.u.cast.mask != 64 ? (1ull << inst.u.cast.mask) - 1 : ~0ull;
                break;
            case OP_BC_GEP1:
            case OP_BC_GEPZ:
                inst.u.three[0] = readNumber(buffer, &offset, len, &ok);
                inst.u.three[1] = readOperand(bcfunc, buffer, &offset, len, &ok);
                inst.u.three[2] = readOperand(bcfunc, buffer, &offset, len, &ok);
                break;
            case OP_BC_GEPN:
                numOp = readFixedNumber(buffer, &offset, len, &ok, 1);
                if (ok) {
                    inst.u.ops.numOps  = numOp + 2;
                    inst.u.ops.opsizes = NULL;
                    inst.u.ops.ops     = cli_calloc(numOp + 2, sizeof(*inst.u.ops.ops));
                    if (!inst.u.ops.ops) {
                        cli_errmsg("Out of memory allocating operands\n");
                        return CL_EMEM;
                    }
                    inst.u.ops.ops[0] = readNumber(buffer, &offset, len, &ok);
                    for (i = 1; i < numOp + 2; i++)
                        inst.u.ops.ops[i] = readOperand(bcfunc, buffer, &offset, len, &ok);
                }
                break;
            case OP_BC_ICMP_EQ:
            case OP_BC_ICMP_NE:
            case OP_BC_ICMP_UGT:
            case OP_BC_ICMP_UGE:
            case OP_BC_ICMP_ULT:
            case OP_BC_ICMP_ULE:
            case OP_BC_ICMP_SGT:
            case OP_BC_ICMP_SGE:
            case OP_BC_ICMP_SLE:
            case OP_BC_ICMP_SLT:
                /* instruction type must be correct before readOperand! */
                inst.type = readNumber(buffer, &offset, len, &ok);
                /* fall-through */
            default:
                numOp = operand_counts[inst.opcode];
                switch (numOp) {
                    case 0:
                        break;
                    case 1:
                        inst.u.unaryop = readOperand(bcfunc, buffer, &offset, len, &ok);
                        break;
                    case 2:
                        inst.u.binop[0] = readOperand(bcfunc, buffer, &offset, len, &ok);
                        inst.u.binop[1] = readOperand(bcfunc, buffer, &offset, len, &ok);
                        break;
                    case 3:
                        inst.u.three[0] = readOperand(bcfunc, buffer, &offset, len, &ok);
                        inst.u.three[1] = readOperand(bcfunc, buffer, &offset, len, &ok);
                        inst.u.three[2] = readOperand(bcfunc, buffer, &offset, len, &ok);
                        break;
                    default:
                        cli_errmsg("Opcode %u with too many operands: %u?\n", inst.opcode, numOp);
                        ok = 0;
                        break;
                }
        }
        if (inst.opcode == OP_BC_STORE) {
            int16_t t = get_optype(bcfunc, inst.u.binop[0]);
            if (t)
                inst.type = t;
        }
        if (inst.opcode == OP_BC_COPY)
            inst.type = get_optype(bcfunc, inst.u.binop[1]);
        if (!ok) {
            cli_errmsg("Invalid instructions or operands\n");
            return CL_EMALFDB;
        }
        if (bcfunc->insn_idx + BB->numInsts >= bcfunc->numInsts) {
            cli_errmsg("More instructions than declared in total: %u > %u!\n",
                       bcfunc->insn_idx + BB->numInsts, bcfunc->numInsts);
            return CL_EMALFDB;
        }
        inst.interp_op = inst.opcode * 5;
        if (inst.type > 1) {
            if (inst.type <= 8)
                inst.interp_op += 1;
            else if (inst.type <= 16)
                inst.interp_op += 2;
            else if (inst.type <= 32)
                inst.interp_op += 3;
            else if (inst.type <= 65)
                inst.interp_op += 4;
            else {
                cli_dbgmsg("unknown inst type: %d\n", inst.type);
            }
        }
        BB->insts[BB->numInsts++] = inst;
    }
    if (bb + 1 == bc->funcs[func].numBB) {
        if (buffer[offset] != 'E') {
            cli_errmsg("Missing basicblock terminator, got: %c\n", buffer[offset]);
            return CL_EMALFDB;
        }
        offset++;
    }
    if (buffer[offset] == 'D') {
        uint32_t num;
        offset += 3;
        if (offset >= len)
            return CL_EMALFDB;
        num = (uint32_t)readNumber(buffer, &offset, len, &ok);
        if (!ok)
            return CL_EMALFDB;
        if (num != bcfunc->numInsts) {
            cli_errmsg("invalid number of dbg nodes, expected: %u, got: %u\n", bcfunc->numInsts, num);
            return CL_EMALFDB;
        }
        bcfunc->dbgnodes = cli_malloc(num * sizeof(*bcfunc->dbgnodes));
        if (!bcfunc->dbgnodes) {
            cli_errmsg("Unable to allocate memory for dbg nodes: %u\n", num * (uint32_t)sizeof(*bcfunc->dbgnodes));
            return CL_EMEM;
        }
        for (i = 0; (uint32_t)i < num; i++) {
            bcfunc->dbgnodes[i] = readNumber(buffer, &offset, len, &ok);
            if (!ok)
                return CL_EMALFDB;
        }
    }
    if (offset != len) {
        cli_errmsg("Trailing garbage in basicblock: %d extra bytes\n",
                   len - offset);
        return CL_EMALFDB;
    }
    bcfunc->numBytes = 0;
    bcfunc->insn_idx += BB->numInsts;
    return CL_SUCCESS;
}

enum parse_state {
    PARSE_BC_TYPES = 0,
    PARSE_BC_APIS,
    PARSE_BC_GLOBALS,
    PARSE_BC_LSIG,
    PARSE_MD_OPT_HEADER,
    PARSE_FUNC_HEADER,
    PARSE_BB,
    PARSE_SKIP
};

struct sigperf_elem {
    const char *bc_name;
    uint64_t usecs;
    unsigned long run_count;
    unsigned long match_count;
};

static int sigelem_comp(const void *a, const void *b)
{
    const struct sigperf_elem *ela = a;
    const struct sigperf_elem *elb = b;
    return elb->usecs / elb->run_count - ela->usecs / ela->run_count;
}

void cli_sigperf_print()
{
    struct sigperf_elem stats[MAX_TRACKED_BC], *elem = stats;
    int i, elems = 0, max_name_len = 0, name_len;

    if (!g_sigid || !g_sigevents) {
        cli_warnmsg("cli_sigperf_print: statistics requested but no bytecodes were loaded!\n");
        return;
    }

    memset(stats, 0, sizeof(stats));
    for (i = 0; i < MAX_TRACKED_BC; i++) {
        union ev_val val;
        uint32_t count;
        const char *name = cli_event_get_name(g_sigevents, i * BC_EVENTS_PER_SIG);
        cli_event_get(g_sigevents, i * BC_EVENTS_PER_SIG, &val, &count);
        if (!count) {
            if (name)
                cli_dbgmsg("No event triggered for %s\n", name);
            continue;
        }
        if (name)
            name_len = (int)strlen(name);
        else
            name_len = 0;
        if (name_len > max_name_len)
            max_name_len = name_len;
        elem->bc_name   = name ? name : "\"noname\"";
        elem->usecs     = val.v_int;
        elem->run_count = count;
        cli_event_get(g_sigevents, i * BC_EVENTS_PER_SIG + 1, &val, &count);
        elem->match_count = count;
        elem++;
        elems++;
    }
    if (max_name_len < (int)strlen("Bytecode name"))
        max_name_len = (int)strlen("Bytecode name");

    cli_qsort(stats, elems, sizeof(struct sigperf_elem), sigelem_comp);

    elem = stats;
    /* name runs matches microsecs avg */
    cli_infomsg(NULL, "%-*s %*s %*s %*s %*s\n", max_name_len, "Bytecode name",
                8, "#runs", 8, "#matches", 12, "usecs total", 9, "usecs avg");
    cli_infomsg(NULL, "%-*s %*s %*s %*s %*s\n", max_name_len, "=============",
                8, "=====", 8, "========", 12, "===========", 9, "=========");
    while (elem->run_count) {
        cli_infomsg(NULL, "%-*s %*lu %*lu %*" PRIu64 " %*.2f\n", max_name_len, elem->bc_name,
                    8, elem->run_count, 8, elem->match_count,
                    12, elem->usecs, 9, (double)elem->usecs / elem->run_count);
        elem++;
    }
}

static void sigperf_events_init(struct cli_bc *bc)
{
    int ret;
    char *bc_name;

    if (!g_sigevents)
        g_sigevents = cli_events_new(MAX_BC_SIGEVENT_ID);

    if (!g_sigevents) {
        cli_errmsg("No memory for events table\n");
        return;
    }

    if (g_sigid > MAX_BC_SIGEVENT_ID - BC_EVENTS_PER_SIG - 1) {
        cli_errmsg("sigperf_events_init: events table full. Increase MAX_TRACKED_BC\n");
        return;
    }

    if (!(bc_name = bc->lsig)) {
        if (!(bc_name = bc->hook_name)) {
            cli_dbgmsg("cli_event_define error for time event id %d\n", bc->sigtime_id);
            return;
        }
    }

    cli_dbgmsg("sigperf_events_init(): adding sig ids starting %u for %s\n", g_sigid, bc_name);

    /* register time event */
    bc->sigtime_id = g_sigid;
    ret            = cli_event_define(g_sigevents, g_sigid++, bc_name, ev_time, multiple_sum);
    if (ret) {
        cli_errmsg("sigperf_events_init: cli_event_define() error for time event id %d\n", bc->sigtime_id);
        bc->sigtime_id = MAX_BC_SIGEVENT_ID + 1;
        return;
    }

    /* register match count */
    bc->sigmatch_id = g_sigid;
    ret             = cli_event_define(g_sigevents, g_sigid++, bc_name, ev_int, multiple_sum);
    if (ret) {
        cli_errmsg("sigperf_events_init: cli_event_define() error for matches event id %d\n", bc->sigmatch_id);
        bc->sigmatch_id = MAX_BC_SIGEVENT_ID + 1;
        return;
    }
}

void cli_sigperf_events_destroy()
{
    cli_events_free(g_sigevents);
}

int cli_bytecode_load(struct cli_bc *bc, FILE *f, struct cli_dbio *dbio, int trust, int sigperf)
{
    unsigned row = 0, current_func = 0, bb = 0;
    char *buffer;
    unsigned linelength = 0;
    char firstbuf[FILEBUFF];
    enum parse_state state;
    int rc, end = 0;

    memset(bc, 0, sizeof(*bc));
    cli_dbgmsg("Loading %s bytecode\n", trust ? "trusted" : "untrusted");
    bc->trusted = trust;
    if (!f && !dbio) {
        cli_errmsg("Unable to load bytecode (null file)\n");
        return CL_ENULLARG;
    }
    if (!cli_dbgets(firstbuf, FILEBUFF, f, dbio)) {
        cli_errmsg("Unable to load bytecode (empty file)\n");
        return CL_EMALFDB;
    }
    cli_chomp(firstbuf);
    rc    = parseHeader(bc, (unsigned char *)firstbuf, &linelength);
    state = PARSE_BC_LSIG;
    if (rc == CL_BREAK) {
        const char *len = strchr(firstbuf, ':');
        bc->state       = bc_skip;
        if (!linelength) {
            linelength = len ? atoi(len + 1) : 4096;
        }
        if (linelength < 4096)
            linelength = 4096;
        cli_dbgmsg("line: %d\n", linelength);
        state = PARSE_SKIP;
        rc    = CL_SUCCESS;
    }
    if (rc != CL_SUCCESS) {
        cli_errmsg("Error at bytecode line %u\n", row);
        return rc;
    }
    buffer = cli_malloc(linelength);
    if (!buffer) {
        cli_errmsg("Out of memory allocating line of length %u\n", linelength);
        return CL_EMEM;
    }
    while (cli_dbgets(buffer, linelength, f, dbio) && !end) {
        cli_chomp(buffer);
        row++;
        switch (state) {
            case PARSE_BC_LSIG:
                rc = parseLSig(bc, buffer);
#if 0
DEAD CODE
		if (rc == CL_BREAK) /* skip */ { //FIXME: parseLSig always returns CL_SUCCESS
		    bc->state = bc_skip;
		    state = PARSE_SKIP;
		    continue;
		}
		if (rc != CL_SUCCESS) { //FIXME: parseLSig always returns CL_SUCCESS
		    cli_errmsg("Error at bytecode line %u\n", row);
		    free(buffer);
		    return rc;
		}
#endif
                state = PARSE_BC_TYPES;
                break;
            case PARSE_BC_TYPES:
                rc = parseTypes(bc, (unsigned char *)buffer);
                if (rc != CL_SUCCESS) {
                    cli_errmsg("Error at bytecode line %u\n", row);
                    free(buffer);
                    return rc;
                }
                state = PARSE_BC_APIS;
                break;
            case PARSE_BC_APIS:
                rc = parseApis(bc, (unsigned char *)buffer);
                if (rc == CL_BREAK) /* skip */ {
                    bc->state = bc_skip;
                    state     = PARSE_SKIP;
                    continue;
                }
                if (rc != CL_SUCCESS) {
                    cli_errmsg("Error at bytecode line %u\n", row);
                    free(buffer);
                    return rc;
                }
                state = PARSE_BC_GLOBALS;
                break;
            case PARSE_BC_GLOBALS:
                rc = parseGlobals(bc, (unsigned char *)buffer);
                if (rc == CL_BREAK) /* skip */ {
                    bc->state = bc_skip;
                    state     = PARSE_SKIP;
                    continue;
                }
                if (rc != CL_SUCCESS) {
                    cli_errmsg("Error at bytecode line %u\n", row);
                    free(buffer);
                    return rc;
                }
                state = PARSE_MD_OPT_HEADER;
                break;
            case PARSE_MD_OPT_HEADER:
                if (buffer[0] == 'D') {
                    rc = parseMD(bc, (unsigned char *)buffer);
                    if (rc != CL_SUCCESS) {
                        cli_errmsg("Error at bytecode line %u\n", row);
                        free(buffer);
                        return rc;
                    }
                    break;
                }
                /* fall-through */
            case PARSE_FUNC_HEADER:
                if (*buffer == 'S') {
                    end = 1;
                    break;
                }
                rc = parseFunctionHeader(bc, current_func, (unsigned char *)buffer);
                if (rc != CL_SUCCESS) {
                    cli_errmsg("Error at bytecode line %u\n", row);
                    free(buffer);
                    return rc;
                }
                bb    = 0;
                state = PARSE_BB;
                break;
            case PARSE_BB:
                rc = parseBB(bc, current_func, bb++, (unsigned char *)buffer);
                if (rc != CL_SUCCESS) {
                    cli_errmsg("Error at bytecode line %u\n", row);
                    free(buffer);
                    return rc;
                }
                if (bb >= bc->funcs[current_func].numBB) {
                    if (bc->funcs[current_func].insn_idx != bc->funcs[current_func].numInsts) {
                        cli_errmsg("Parsed different number of instructions than declared: %u != %u\n",
                                   bc->funcs[current_func].insn_idx, bc->funcs[current_func].numInsts);
                        free(buffer);
                        return CL_EMALFDB;
                    }
                    cli_dbgmsg("Parsed %u BBs, %u instructions\n",
                               bb, bc->funcs[current_func].numInsts);
                    state = PARSE_FUNC_HEADER;
                    current_func++;
                }
                break;
            case PARSE_SKIP:
                /* stop at S (source code), readdb.c knows how to skip this one
                 * */
                if (buffer[0] == 'S')
                    end = 1;
                /* noop parse, but we need to use dbgets with dynamic buffer,
                 * otherwise we get 'Line too long for provided buffer' */
                break;
        }
    }
    free(buffer);
    cli_dbgmsg("Parsed %d functions\n", current_func);
    if (sigperf)
        sigperf_events_init(bc);
    if (current_func != bc->num_func && bc->state != bc_skip) {
        cli_errmsg("Loaded less functions than declared: %u vs. %u\n",
                   current_func, bc->num_func);
        return CL_EMALFDB;
    }
    return CL_SUCCESS;
}

static struct {
    enum bc_events id;
    const char *name;
    enum ev_type type;
    enum multiple_handling multiple;
} bc_events[] = {
    {BCEV_VIRUSNAME, "virusname", ev_string, multiple_last},
    {BCEV_EXEC_RETURNVALUE, "returnvalue", ev_int, multiple_last},
    {BCEV_WRITE, "bcapi_write", ev_data_fast, multiple_sum},
    {BCEV_OFFSET, "read offset", ev_int, multiple_sum},
    {BCEV_READ, "read data", ev_data_fast, multiple_sum},
    //{BCEV_READ, "read data", ev_data, multiple_concat},
    {BCEV_DBG_STR, "debug message", ev_data_fast, multiple_sum},
    {BCEV_DBG_INT, "debug int", ev_int, multiple_sum},
    {BCEV_MEM_1, "memmem 1", ev_data_fast, multiple_sum},
    {BCEV_MEM_2, "memmem 2", ev_data_fast, multiple_sum},
    {BCEV_FIND, "find", ev_data_fast, multiple_sum},
    {BCEV_EXTRACTED, "extracted files", ev_int, multiple_sum},
    {BCEV_READ_ERR, "read errors", ev_int, multiple_sum},
    {BCEV_DISASM_FAIL, "disasm fails", ev_int, multiple_sum},
    {BCEV_EXEC_TIME, "bytecode execute", ev_time, multiple_sum}};

static int register_events(cli_events_t *ev)
{
    size_t i;
    for (i = 0; i < sizeof(bc_events) / sizeof(bc_events[0]); i++) {
        if (cli_event_define(ev, bc_events[i].id, bc_events[i].name, bc_events[i].type,
                             bc_events[i].multiple) == -1)
            return -1;
    }
    return 0;
}

int cli_bytecode_run(const struct cli_all_bc *bcs, const struct cli_bc *bc, struct cli_bc_ctx *ctx)
{
    int ret = CL_SUCCESS;
    struct cli_bc_inst inst;
    struct cli_bc_func func;
    cli_events_t *jit_ev = NULL, *interp_ev = NULL;

    int test_mode = 0;
    cli_ctx *cctx = (cli_ctx *)ctx->ctx;

    if (!ctx || !ctx->bc || !ctx->func)
        return CL_ENULLARG;
    if (ctx->numParams && (!ctx->values || !ctx->operands))
        return CL_ENULLARG;

    if (cctx && cctx->engine->bytecode_mode == CL_BYTECODE_MODE_TEST)
        test_mode = 1;

    if (bc->state == bc_loaded) {
        cli_errmsg("bytecode has to be prepared either for interpreter or JIT!\n");
        return CL_EARG;
    }
    if (bc->state == bc_disabled) {
        cli_dbgmsg("bytecode triggered but running bytecodes is disabled\n");
        return CL_SUCCESS;
    }
    if (cctx)
        cli_event_time_start(cctx->perf, PERFT_BYTECODE);
    ctx->env = &bcs->env;
    context_safe(ctx);
    if (test_mode) {
        jit_ev    = cli_events_new(BCEV_LASTEVENT);
        interp_ev = cli_events_new(BCEV_LASTEVENT);
        if (!jit_ev || !interp_ev) {
            cli_events_free(jit_ev);
            cli_events_free(interp_ev);
            return CL_EMEM;
        }
        if (register_events(jit_ev) == -1 ||
            register_events(interp_ev) == -1) {
            cli_events_free(jit_ev);
            cli_events_free(interp_ev);
            return CL_EBYTECODE_TESTFAIL;
        }
    }
    cli_event_time_start(g_sigevents, bc->sigtime_id);
    if (bc->state == bc_interp || test_mode) {
        ctx->bc_events = interp_ev;
        memset(&func, 0, sizeof(func));
        func.numInsts     = 1;
        func.numValues    = 1;
        func.numConstants = 0;
        func.numBytes     = ctx->bytes;
        memset(ctx->values + ctx->bytes - 8, 0, 8);

        inst.opcode        = OP_BC_CALL_DIRECT;
        inst.interp_op     = OP_BC_CALL_DIRECT * 5;
        inst.dest          = func.numArgs;
        inst.type          = 0;
        inst.u.ops.numOps  = ctx->numParams;
        inst.u.ops.funcid  = ctx->funcid;
        inst.u.ops.ops     = ctx->operands;
        inst.u.ops.opsizes = ctx->opsizes;
        cli_dbgmsg("Bytecode %u: executing in interpreter mode\n", bc->id);

        ctx->on_jit = 0;

        cli_event_time_start(interp_ev, BCEV_EXEC_TIME);
        ret = cli_vm_execute(ctx->bc, ctx, &func, &inst);
        cli_event_time_stop(interp_ev, BCEV_EXEC_TIME);

        cli_event_int(interp_ev, BCEV_EXEC_RETURNVALUE, ret);
        cli_event_string(interp_ev, BCEV_VIRUSNAME, ctx->virname);

        /* need to be called here to catch any extracted but not yet scanned files */
        if (ctx->outfd && (ret != CL_VIRUS || cctx->options->general & CL_SCAN_GENERAL_ALLMATCHES))
            cli_bcapi_extract_new(ctx, -1);
    }
    if (bc->state == bc_jit || test_mode) {
        if (test_mode) {
            ctx->off = 0;
        }
        ctx->bc_events = jit_ev;
        cli_dbgmsg("Bytecode %u: executing in JIT mode\n", bc->id);

        ctx->on_jit = 1;
        cli_event_time_start(jit_ev, BCEV_EXEC_TIME);
        ret = cli_vm_execute_jit(bcs, ctx, &bc->funcs[ctx->funcid]);
        cli_event_time_stop(jit_ev, BCEV_EXEC_TIME);

        cli_event_int(jit_ev, BCEV_EXEC_RETURNVALUE, ret);
        cli_event_string(jit_ev, BCEV_VIRUSNAME, ctx->virname);

        /* need to be called here to catch any extracted but not yet scanned files */
        if (ctx->outfd && (ret != CL_VIRUS || cctx->options->general & CL_SCAN_GENERAL_ALLMATCHES))
            cli_bcapi_extract_new(ctx, -1);
    }
    cli_event_time_stop(g_sigevents, bc->sigtime_id);
    if (ctx->virname)
        cli_event_count(g_sigevents, bc->sigmatch_id);

    if (test_mode) {
        unsigned interp_errors = cli_event_errors(interp_ev);
        unsigned jit_errors    = cli_event_errors(jit_ev);
        unsigned interp_warns = 0, jit_warns = 0;
        int ok = 1;
        enum bc_events evid;

        if (interp_errors || jit_errors) {
            cli_infomsg(cctx, "bytecode %d encountered %u JIT and %u interpreter errors\n",
                        bc->id, interp_errors, jit_errors);
            ok = 0;
        }
        if (!ctx->no_diff && cli_event_diff_all(interp_ev, jit_ev, NULL)) {
            cli_infomsg(cctx, "bytecode %d execution different with JIT and interpreter, see --debug for details\n",
                        bc->id);
            ok = 0;
        }
        for (evid = BCEV_API_WARN_BEGIN + 1; evid < BCEV_API_WARN_END; evid++) {
            union ev_val v;
            uint32_t count = 0;
            cli_event_get(interp_ev, evid, &v, &count);
            interp_warns += count;
            count = 0;
            cli_event_get(jit_ev, evid, &v, &count);
            jit_warns += count;
        }
        if (interp_warns || jit_warns) {
            cli_infomsg(cctx, "bytecode %d encountered %u JIT and %u interpreter warnings\n",
                        bc->id, interp_warns, jit_warns);
            ok = 0;
        }
        /*cli_event_debug(jit_ev, BCEV_EXEC_TIME);
        cli_event_debug(interp_ev, BCEV_EXEC_TIME);
        cli_event_debug(g_sigevents, bc->sigtime_id);*/
        if (!ok) {
            cli_events_free(jit_ev);
            cli_events_free(interp_ev);
            return CL_EBYTECODE_TESTFAIL;
        }
    }
    cli_events_free(jit_ev);
    cli_events_free(interp_ev);
    if (cctx)
        cli_event_time_stop(cctx->perf, PERFT_BYTECODE);
    return ret;
}

uint64_t cli_bytecode_context_getresult_int(struct cli_bc_ctx *ctx)
{
    return *(uint32_t *)ctx->values; /*XXX*/
}

void cli_bytecode_destroy(struct cli_bc *bc)
{
    unsigned i, j, k;
    free(bc->metadata.compiler);
    free(bc->metadata.sigmaker);

    if (bc->funcs) {
        for (i = 0; i < bc->num_func; i++) {
            struct cli_bc_func *f = &bc->funcs[i];
            if (!f)
                continue;
            free(f->types);

            for (j = 0; j < f->numBB; j++) {
                struct cli_bc_bb *BB = &f->BB[j];
                for (k = 0; k < BB->numInsts; k++) {
                    struct cli_bc_inst *ii = &BB->insts[k];
                    if (operand_counts[ii->opcode] > 3 ||
                        ii->opcode == OP_BC_CALL_DIRECT || ii->opcode == OP_BC_CALL_API) {
                        free(ii->u.ops.ops);
                        free(ii->u.ops.opsizes);
                    }
                }
            }
            free(f->BB);
            free(f->allinsts);
            free(f->constants);
        }
        free(bc->funcs);
    }
    if (bc->types) {
        for (i = NUM_STATIC_TYPES; i < bc->num_types; i++) {
            if (bc->types[i].containedTypes)
                free(bc->types[i].containedTypes);
        }
        free(bc->types);
    }

    if (bc->globals) {
        for (i = 0; i < bc->num_globals; i++) {
            free(bc->globals[i]);
        }
        free(bc->globals);
    }
    if (bc->dbgnodes) {
        for (i = 0; i < bc->dbgnode_cnt; i++) {
            for (j = 0; j < bc->dbgnodes[i].numelements; j++) {
                struct cli_bc_dbgnode_element *el = &bc->dbgnodes[i].elements[j];
                if (el && el->string)
                    free(el->string);
            }
        }
        free(bc->dbgnodes);
    }
    free(bc->globaltys);
    if (bc->uses_apis)
        cli_bitset_free(bc->uses_apis);
    free(bc->lsig);
    free(bc->hook_name);
    free(bc->globalBytes);
    memset(bc, 0, sizeof(*bc));
}

#define MAP(val)                                                                                                            \
    do {                                                                                                                    \
        operand_t o = val;                                                                                                  \
        if (o & 0x80000000) {                                                                                               \
            o &= 0x7fffffff;                                                                                                \
            if (o > bc->num_globals) {                                                                                      \
                cli_errmsg("bytecode: global out of range: %u > %u, for instruction %u in function %u\n",                   \
                           o, (unsigned)bc->num_globals, j, i);                                                             \
                free(map);                                                                                                  \
                free(gmap);                                                                                                 \
                return CL_EBYTECODE;                                                                                        \
            }                                                                                                               \
            val = 0x80000000 | gmap[o];                                                                                     \
            break;                                                                                                          \
        }                                                                                                                   \
        if (o >= totValues) {                                                                                               \
            cli_errmsg("bytecode: operand out of range: %u > %u, for instruction %u in function %u\n", o, totValues, j, i); \
            free(map);                                                                                                      \
            free(gmap);                                                                                                     \
            return CL_EBYTECODE;                                                                                            \
        }                                                                                                                   \
        val = map[o];                                                                                                       \
    } while (0)

#define MAPPTR(val)                                                   \
    {                                                                 \
        if ((val < bcfunc->numValues) && bcfunc->types[val] & 0x8000) \
            val = map[val] | 0x40000000;                              \
        else                                                          \
            MAP(val);                                                 \
    }

static inline int64_t ptr_compose(int32_t id, uint32_t offset)
{
    uint64_t i = id;
    return (i << 32) | offset;
}

static inline int get_geptypesize(const struct cli_bc *bc, uint16_t tid)
{
    const struct cli_bc_type *ty;
    if (tid >= bc->num_types + 65) {
        cli_errmsg("bytecode: typeid out of range %u >= %u\n", tid, bc->num_types);
        return -1;
    }
    if (tid <= 64) {
        cli_errmsg("bytecode: invalid type for gep (%u)\n", tid);
        return -1;
    }
    ty = &bc->types[tid - 65];
    if (ty->kind != DPointerType) {
        cli_errmsg("bytecode: invalid gep type, must be pointer: %u\n", tid);
        return -1;
    }
    return typesize(bc, ty->containedTypes[0]);
}

static int calc_gepz(struct cli_bc *bc, struct cli_bc_func *func, uint16_t tid, operand_t op)
{
    unsigned off = 0, i;
    uint32_t *gepoff;
    const struct cli_bc_type *ty;
    if (tid >= bc->num_types + 65) {
        cli_errmsg("bytecode: typeid out of range %u >= %u\n", tid, bc->num_types);
        return -1;
    }
    if (tid <= 65) {
        cli_errmsg("bytecode: invalid type for gep (%u)\n", tid);
        return -1;
    }
    ty = &bc->types[tid - 65];
    if (ty->kind != DPointerType || ty->containedTypes[0] < 65) {
        cli_errmsg("bytecode: invalid gep type, must be pointer to nonint: %u\n", tid);
        return -1;
    }
    ty = &bc->types[ty->containedTypes[0] - 65];
    if (ty->kind != DStructType && ty->kind != DPackedStructType)
        return 0;
    gepoff = (uint32_t *)&func->constants[op - func->numValues];
    if (*gepoff >= ty->numElements) {
        cli_errmsg("bytecode: gep offset out of range: %d >= %d\n", (uint32_t)*gepoff, ty->numElements);
        return -1;
    }
    for (i = 0; i < *gepoff; i++) {
        off += typesize(bc, ty->containedTypes[i]);
    }
    *gepoff = off;
    return 1;
}

static int cli_bytecode_prepare_interpreter(struct cli_bc *bc)
{
    unsigned i, j, k;
    uint64_t *gmap;
    unsigned bcglobalid = cli_apicall_maxglobal - _FIRST_GLOBAL + 2;
    int ret             = CL_SUCCESS;
    bc->numGlobalBytes  = 0;
    gmap                = cli_malloc(bc->num_globals * sizeof(*gmap));
    if (!gmap) {
        cli_errmsg("interpreter: Unable to allocate memory for global map: %zu\n", bc->num_globals * sizeof(*gmap));
        return CL_EMEM;
    }
    for (j = 0; j < bc->num_globals; j++) {
        uint16_t ty    = bc->globaltys[j];
        unsigned align = typealign(bc, ty);
        assert(align);
        bc->numGlobalBytes = (bc->numGlobalBytes + align - 1) & (~(align - 1));
        gmap[j]            = bc->numGlobalBytes;
        bc->numGlobalBytes += typesize(bc, ty);
    }
    if (bc->numGlobalBytes) {
        bc->globalBytes = cli_calloc(1, bc->numGlobalBytes);
        if (!bc->globalBytes) {
            cli_errmsg("interpreter: Unable to allocate memory for globalBytes: %u\n", bc->numGlobalBytes);
            free(gmap);
            return CL_EMEM;
        }
    } else
        bc->globalBytes = NULL;

    for (j = 0; j < bc->num_globals; j++) {
        struct cli_bc_type *ty;
        if (bc->globaltys[j] < 65)
            continue;
        ty = &bc->types[bc->globaltys[j] - 65];
        switch (ty->kind) {
            case DPointerType: {
                uint64_t ptr;
                if (bc->globals[j][1] >= _FIRST_GLOBAL) {
                    ptr = ptr_compose(bc->globals[j][1] - _FIRST_GLOBAL + 1,
                                      bc->globals[j][0]);
                } else {
                    if (bc->globals[j][1] > bc->num_globals)
                        continue;
                    ptr = ptr_compose(bcglobalid,
                                      gmap[bc->globals[j][1]] + bc->globals[j][0]);
                }
                *(uint64_t *)&bc->globalBytes[gmap[j]] = ptr;
                break;
            }
            case DArrayType: {
                unsigned elsize, i, off = gmap[j];
                /* TODO: support other than ints in arrays */
                elsize = typesize(bc, ty->containedTypes[0]);
                switch (elsize) {
                    case 1:
                        for (i = 0; i < ty->numElements; i++)
                            bc->globalBytes[off + i] = bc->globals[j][i];
                        break;
                    case 2:
                        for (i = 0; i < ty->numElements; i++)
                            *(uint16_t *)&bc->globalBytes[off + i * 2] = bc->globals[j][i];
                        break;
                    case 4:
                        for (i = 0; i < ty->numElements; i++)
                            *(uint32_t *)&bc->globalBytes[off + i * 4] = bc->globals[j][i];
                        break;
                    case 8:
                        for (i = 0; i < ty->numElements; i++)
                            *(uint64_t *)&bc->globalBytes[off + i * 8] = bc->globals[j][i];
                        break;
                    default:
                        cli_dbgmsg("interpreter: unsupported elsize: %u\n", elsize);
                }
                break;
            }
            default:
                /*TODO*/
                if (!bc->globals[j][1])
                    continue; /* null */
                break;
        }
    }

    for (i = 0; i < bc->num_func && ret == CL_SUCCESS; i++) {
        struct cli_bc_func *bcfunc = &bc->funcs[i];
        unsigned totValues         = bcfunc->numValues + bcfunc->numConstants + bc->num_globals;
        unsigned *map              = cli_malloc(sizeof(*map) * (size_t)totValues);
        if (!map) {
            cli_errmsg("interpreter: Unable to allocate memory for map: %zu\n", sizeof(*map) * (size_t)totValues);
            free(gmap);
            return CL_EMEM;
        }
        bcfunc->numBytes = 0;
        for (j = 0; j < bcfunc->numValues; j++) {
            uint16_t ty = bcfunc->types[j];
            unsigned align;
            align = typealign(bc, ty);
            assert(!ty || typesize(bc, ty));
            assert(align);
            bcfunc->numBytes = (bcfunc->numBytes + align - 1) & (~(align - 1));
            map[j]           = bcfunc->numBytes;
            /* printf("%d -> %d, %u\n", j, map[j], typesize(bc, ty)); */
            bcfunc->numBytes += typesize(bc, ty);
            /* TODO: don't allow size 0, it is always a bug! */
        }
        bcfunc->numBytes = (bcfunc->numBytes + 7) & ~7;
        for (j = 0; j < bcfunc->numConstants; j++) {
            map[bcfunc->numValues + j] = bcfunc->numBytes;
            bcfunc->numBytes += 8;
        }
        for (j = 0; j < bcfunc->numInsts && ret == CL_SUCCESS; j++) {
            struct cli_bc_inst *inst = &bcfunc->allinsts[j];
            inst->dest               = map[inst->dest];
            switch (inst->opcode) {
                case OP_BC_ADD:
                case OP_BC_SUB:
                case OP_BC_MUL:
                case OP_BC_UDIV:
                case OP_BC_SDIV:
                case OP_BC_UREM:
                case OP_BC_SREM:
                case OP_BC_SHL:
                case OP_BC_LSHR:
                case OP_BC_ASHR:
                case OP_BC_AND:
                case OP_BC_OR:
                case OP_BC_XOR:
                case OP_BC_ICMP_EQ:
                case OP_BC_ICMP_NE:
                case OP_BC_ICMP_UGT:
                case OP_BC_ICMP_UGE:
                case OP_BC_ICMP_ULT:
                case OP_BC_ICMP_ULE:
                case OP_BC_ICMP_SGT:
                case OP_BC_ICMP_SGE:
                case OP_BC_ICMP_SLT:
                case OP_BC_ICMP_SLE:
                case OP_BC_COPY:
                case OP_BC_STORE:
                    MAP(inst->u.binop[0]);
                    MAP(inst->u.binop[1]);
                    break;
                case OP_BC_SEXT:
                case OP_BC_ZEXT:
                case OP_BC_TRUNC:
                    MAP(inst->u.cast.source);
                    break;
                case OP_BC_BRANCH:
                    MAP(inst->u.branch.condition);
                    break;
                case OP_BC_JMP:
                    break;
                case OP_BC_RET:
                    MAP(inst->u.unaryop);
                    break;
                case OP_BC_SELECT:
                    MAP(inst->u.three[0]);
                    MAP(inst->u.three[1]);
                    MAP(inst->u.three[2]);
                    break;
                case OP_BC_CALL_API: /* fall-through */
                case OP_BC_CALL_DIRECT: {
                    struct cli_bc_func *target = NULL;
                    if (inst->opcode == OP_BC_CALL_DIRECT) {
                        target = &bc->funcs[inst->u.ops.funcid];
                        if (inst->u.ops.funcid > bc->num_func) {
                            cli_errmsg("bytecode: called function out of range: %u > %u\n", inst->u.ops.funcid, bc->num_func);
                            ret = CL_EBYTECODE;
                        } else if (inst->u.ops.numOps != target->numArgs) {
                            cli_errmsg("bytecode: call operands don't match function prototype\n");
                            ret = CL_EBYTECODE;
                        }
                    } else {
                        /* APIs have at most 2 parameters always */
                        if (inst->u.ops.numOps > 5) {
                            cli_errmsg("bytecode: call operands don't match function prototype\n");
                            ret = CL_EBYTECODE;
                        }
                    }
                    if (ret != CL_SUCCESS)
                        break;
                    if (inst->u.ops.numOps > 0) {
                        inst->u.ops.opsizes = cli_malloc(sizeof(*inst->u.ops.opsizes) * inst->u.ops.numOps);
                        if (!inst->u.ops.opsizes) {
                            cli_errmsg("Out of memory when allocating operand sizes\n");
                            ret = CL_EMEM;
                            break;
                        }
                    } else {
                        inst->u.ops.opsizes = NULL;
                        break;
                    }
                    for (k = 0; k < inst->u.ops.numOps; k++) {
                        MAPPTR(inst->u.ops.ops[k]);
                        if (inst->opcode == OP_BC_CALL_DIRECT)
                            inst->u.ops.opsizes[k] = typesize(bc, target->types[k]);
                        else
                            inst->u.ops.opsizes[k] = 32; /*XXX*/
                    }
                    break;
                }
                case OP_BC_LOAD:
                    MAPPTR(inst->u.unaryop);
                    break;
                case OP_BC_GEP1:
                    if (inst->u.three[1] & 0x80000000 ||
                        bcfunc->types[inst->u.binop[1]] & 0x8000) {
                        cli_errmsg("bytecode: gep1 of alloca is not allowed\n");
                        ret = CL_EBYTECODE;
                    }
                    if (ret != CL_SUCCESS)
                        break;
                    MAP(inst->u.three[1]);
                    MAP(inst->u.three[2]);
                    inst->u.three[0] = get_geptypesize(bc, inst->u.three[0]);
                    if ((int)(inst->u.three[0]) == -1)
                        ret = CL_EBYTECODE;
                    break;
                case OP_BC_GEPZ:
                    /*three[0] is the type*/
                    if (inst->u.three[1] & 0x80000000 ||
                        bcfunc->types[inst->u.three[1]] & 0x8000)
                        inst->interp_op = 5 * (inst->interp_op / 5);
                    else
                        inst->interp_op = 5 * (inst->interp_op / 5) + 3;
                    MAP(inst->u.three[1]);
                    if (calc_gepz(bc, bcfunc, inst->u.three[0], inst->u.three[2]) == -1)
                        ret = CL_EBYTECODE;
                    if (ret == CL_SUCCESS)
                        MAP(inst->u.three[2]);
                    break;
                    /*		case OP_BC_GEPN:
                    *TODO
                    break;*/
                case OP_BC_MEMSET:
                case OP_BC_MEMCPY:
                case OP_BC_MEMMOVE:
                case OP_BC_MEMCMP:
                    MAPPTR(inst->u.three[0]);
                    MAPPTR(inst->u.three[1]);
                    MAP(inst->u.three[2]);
                    break;
                case OP_BC_RET_VOID:
                case OP_BC_ISBIGENDIAN:
                case OP_BC_ABORT:
                    /* no operands */
                    break;
                case OP_BC_BSWAP16:
                case OP_BC_BSWAP32:
                case OP_BC_BSWAP64:
                    MAP(inst->u.unaryop);
                    break;
                case OP_BC_PTRDIFF32:
                    MAPPTR(inst->u.binop[0]);
                    MAPPTR(inst->u.binop[1]);
                    break;
                case OP_BC_PTRTOINT64:
                    MAPPTR(inst->u.unaryop);
                    break;
                default:
                    cli_warnmsg("Bytecode: unhandled opcode: %d\n", inst->opcode);
                    ret = CL_EBYTECODE;
            }
        }
        if (map)
            free(map);
    }
    free(gmap);
    bc->state = bc_interp;
    return ret;
}

static int add_selfcheck(struct cli_all_bc *bcs)
{
    struct cli_bc_func *func;
    struct cli_bc_inst *inst;
    struct cli_bc *bc;

    bcs->all_bcs = cli_realloc2(bcs->all_bcs, sizeof(*bcs->all_bcs) * (bcs->count + 1));
    if (!bcs->all_bcs) {
        cli_errmsg("cli_loadcbc: Can't allocate memory for bytecode entry\n");
        return CL_EMEM;
    }
    bc = &bcs->all_bcs[bcs->count++];
    memset(bc, 0, sizeof(*bc));

    bc->trusted     = 1;
    bc->num_globals = 1;
    bc->globals     = cli_calloc(1, sizeof(*bc->globals));
    if (!bc->globals) {
        cli_errmsg("Failed to allocate memory for globals\n");
        return CL_EMEM;
    }
    bc->globals[0] = cli_calloc(1, sizeof(*bc->globals[0]));
    if (!bc->globals[0]) {
        cli_errmsg("Failed to allocate memory for globals\n");
        return CL_EMEM;
    }
    bc->globaltys = cli_calloc(1, sizeof(*bc->globaltys));
    if (!bc->globaltys) {
        cli_errmsg("Failed to allocate memory for globaltypes\n");
        return CL_EMEM;
    }
    bc->globaltys[0] = 32;
    *bc->globals[0]  = 0;
    bc->id           = ~0;
    bc->kind         = 0;
    bc->num_types    = 5;
    bc->num_func     = 1;
    bc->funcs        = cli_calloc(1, sizeof(*bc->funcs));
    if (!bc->funcs) {
        cli_errmsg("Failed to allocate memory for func\n");
        return CL_EMEM;
    }
    func               = bc->funcs;
    func->numInsts     = 2;
    func->numLocals    = 1;
    func->numValues    = 1;
    func->numConstants = 1;
    func->numBB        = 1;
    func->returnType   = 32;
    func->types        = cli_calloc(1, sizeof(*func->types));
    if (!func->types) {
        cli_errmsg("Failed to allocate memory for types\n");
        return CL_EMEM;
    }
    func->types[0] = 32;
    func->BB       = cli_calloc(1, sizeof(*func->BB));
    if (!func->BB) {
        cli_errmsg("Failed to allocate memory for BB\n");
        return CL_EMEM;
    }
    func->allinsts = cli_calloc(2, sizeof(*func->allinsts));
    if (!func->allinsts) {
        cli_errmsg("Failed to allocate memory for insts\n");
        return CL_EMEM;
    }
    func->BB->numInsts = 2;
    func->BB->insts    = func->allinsts;
    func->constants    = cli_calloc(1, sizeof(*func->constants));
    if (!func->constants) {
        cli_errmsg("Failed to allocate memory for constants\n");
        return CL_EMEM;
    }
    func->constants[0] = 0xf00d;
    inst               = func->allinsts;

    inst->opcode        = OP_BC_CALL_API;
    inst->u.ops.numOps  = 1;
    inst->u.ops.opsizes = NULL;
    inst->u.ops.ops     = cli_calloc(1, sizeof(*inst->u.ops.ops));
    if (!inst->u.ops.ops) {
        cli_errmsg("Failed to allocate memory for instructions\n");
        return CL_EMEM;
    }
    inst->u.ops.ops[0] = 1;
    inst->u.ops.funcid = 18; /* test2 */
    inst->dest         = 0;
    inst->type         = 32;
    inst->interp_op    = inst->opcode * 5 + 3;

    inst            = &func->allinsts[1];
    inst->opcode    = OP_BC_RET;
    inst->type      = 32;
    inst->u.unaryop = 0;
    inst->interp_op = inst->opcode * 5;

    bc->state = bc_loaded;
    return 0;
}

static int run_selfcheck(struct cli_all_bc *bcs)
{
    struct cli_bc_ctx *ctx;
    struct cli_bc *bc = &bcs->all_bcs[bcs->count - 1];
    int rc;
    if (bc->state != bc_jit && bc->state != bc_interp) {
        cli_errmsg("Failed to prepare selfcheck bytecode\n");
        return CL_EBYTECODE;
    }
    ctx = cli_bytecode_context_alloc();
    if (!ctx) {
        cli_errmsg("Failed to allocate bytecode context\n");
        return CL_EMEM;
    }
    cli_bytecode_context_setfuncid(ctx, bc, 0);

    cli_dbgmsg("bytecode self test running\n");
    ctx->bytecode_timeout = 0;
    rc                    = cli_bytecode_run(bcs, bc, ctx);
    cli_bytecode_context_destroy(ctx);
    if (rc != CL_SUCCESS) {
        cli_errmsg("bytecode self test failed: %s\n",
                   cl_strerror(rc));
    } else {
        cli_dbgmsg("bytecode self test succeeded\n");
    }
    return rc;
}

static int selfcheck(int jit, struct cli_bcengine *engine)
{
    struct cli_all_bc bcs;
    int rc;

    memset(&bcs, 0, sizeof(bcs));
    bcs.all_bcs = NULL;
    bcs.count   = 0;
    bcs.engine  = engine;
    rc          = add_selfcheck(&bcs);
    if (rc == CL_SUCCESS) {
        if (jit) {
            if (!bcs.engine) {
                cli_dbgmsg("bytecode: JIT disabled\n");
                rc = CL_BREAK; /* no JIT - not fatal */
            } else {
                rc = cli_bytecode_prepare_jit(&bcs);
            }
        } else {
            rc = cli_bytecode_prepare_interpreter(bcs.all_bcs);
        }
        if (rc == CL_SUCCESS)
            rc = run_selfcheck(&bcs);
        if (rc == CL_BREAK)
            rc = CL_SUCCESS;
    }
    cli_bytecode_destroy(bcs.all_bcs);
    free(bcs.all_bcs);
    cli_bytecode_done_jit(&bcs, 1);
    if (rc != CL_SUCCESS) {
        cli_errmsg("Bytecode: failed to run selfcheck in %s mode: %s\n",
                   jit ? "JIT" : "interpreter", cl_strerror(rc));
    }
    return rc;
}

static int set_mode(struct cl_engine *engine, enum bytecode_mode mode)
{
    if (engine->bytecode_mode == mode)
        return 0;
    if (engine->bytecode_mode == CL_BYTECODE_MODE_OFF) {
        cli_errmsg("bytecode: already turned off, can't turn it on again!\n");
        return -1;
    }
    cli_dbgmsg("Bytecode: mode changed to %d\n", mode);
    if (engine->bytecode_mode == CL_BYTECODE_MODE_TEST) {
        if (mode == CL_BYTECODE_MODE_OFF || have_clamjit()) {
            cli_errmsg("bytecode: in test mode but JIT/bytecode is about to be disabled: %d\n", mode);
            engine->bytecode_mode = mode;
            return -1;
        }
        return 0;
    }
    if (engine->bytecode_mode == CL_BYTECODE_MODE_JIT) {
        cli_errmsg("bytecode: in JIT mode but JIT is about to be disabled: %d\n", mode);
        engine->bytecode_mode = mode;
        return -1;
    }
    engine->bytecode_mode = mode;
    return 0;
}

/* runs the first bytecode of the specified kind, or the builtin one if no
 * bytecode of that kind is loaded */
static int run_builtin_or_loaded(struct cli_all_bc *bcs, uint8_t kind, const char *builtin_cbc, struct cli_bc_ctx *ctx, const char *desc)
{
    unsigned i, builtin = 0, rc = 0;
    struct cli_bc *bc = NULL;

    for (i = 0; i < bcs->count; i++) {
        bc = &bcs->all_bcs[i];
        if (bc->kind == kind)
            break;
    }
    if (i == bcs->count)
        bc = NULL;
    if (!bc) {
        /* no loaded bytecode found, load the builtin one! */
        struct cli_dbio dbio;
        bc = cli_calloc(1, sizeof(*bc));
        if (!bc) {
            cli_errmsg("Out of memory allocating bytecode\n");
            return CL_EMEM;
        }
        builtin = 1;

        memset(&dbio, 0, sizeof(dbio));
        dbio.usebuf = 1;
        dbio.bufpt = dbio.buf = (char *)builtin_cbc;
        dbio.bufsize          = strlen(builtin_cbc) + 1;
        if (!dbio.bufsize || dbio.bufpt[dbio.bufsize - 2] != '\n') {
            cli_errmsg("Invalid builtin bytecode: missing terminator\n");
            free(bc);
            return CL_EMALFDB;
        }

        rc = cli_bytecode_load(bc, NULL, &dbio, 1, 0);
        if (rc) {
            cli_errmsg("Failed to load builtin %s bytecode\n", desc);
            free(bc);
            return rc;
        }
    }
    rc = cli_bytecode_prepare_interpreter(bc);
    if (rc) {
        cli_errmsg("Failed to prepare %s %s bytecode for interpreter: %s\n",
                   builtin ? "builtin" : "loaded", desc, cl_strerror(rc));
    }
    if (bc->state != bc_interp) {
        cli_errmsg("Failed to prepare %s %s bytecode for interpreter\n",
                   builtin ? "builtin" : "loaded", desc);
        rc = CL_EMALFDB;
    }
    if (!rc) {
        cli_bytecode_context_setfuncid(ctx, bc, 0);
        cli_dbgmsg("Bytecode: %s running (%s)\n", desc,
                   builtin ? "builtin" : "loaded");
        rc = cli_bytecode_run(bcs, bc, ctx);
    }
    if (rc) {
        cli_errmsg("Failed to execute %s %s bytecode: %s\n", builtin ? "builtin" : "loaded",
                   desc, cl_strerror(rc));
    }
    if (builtin) {
        cli_bytecode_destroy(bc);
        free(bc);
    }
    return rc;
}

int cli_bytecode_prepare2(struct cl_engine *engine, struct cli_all_bc *bcs, unsigned dconfmask)
{
    unsigned i, interp = 0, jitok = 0, jitcount = 0;
    int rc;
    struct cli_bc_ctx *ctx;

    if (!bcs->count) {
        cli_dbgmsg("No bytecodes loaded, not running builtin test\n");
        return CL_SUCCESS;
    }

    engine->bytecode_mode = CL_BYTECODE_MODE_AUTO;
    cli_detect_environment(&bcs->env);
    switch (bcs->env.arch) {
        case arch_i386:
        case arch_x86_64:
            if (!(dconfmask & BYTECODE_JIT_X86)) {
                cli_dbgmsg("Bytecode: disabled on X86 via DCONF\n");
                if (set_mode(engine, CL_BYTECODE_MODE_INTERPRETER) == -1)
                    return CL_EBYTECODE_TESTFAIL;
            }
            break;
        case arch_ppc32:
        case arch_ppc64:
            if (!(dconfmask & BYTECODE_JIT_PPC)) {
                cli_dbgmsg("Bytecode: disabled on PPC via DCONF\n");
                if (set_mode(engine, CL_BYTECODE_MODE_INTERPRETER) == -1)
                    return CL_EBYTECODE_TESTFAIL;
            }
            break;
        case arch_arm:
            if (!(dconfmask & BYTECODE_JIT_ARM)) {
                cli_dbgmsg("Bytecode: disabled on ARM via DCONF\n");
                if (set_mode(engine, CL_BYTECODE_MODE_INTERPRETER) == -1)
                    return CL_EBYTECODE_TESTFAIL;
            }
            break;
        default:
            cli_dbgmsg("Bytecode: JIT not supported on this architecture, falling back\n");
            if (set_mode(engine, CL_BYTECODE_MODE_INTERPRETER) == -1)
                return CL_EBYTECODE_TESTFAIL;
            break;
    }
    cli_dbgmsg("Bytecode: mode is %d\n", engine->bytecode_mode);

    ctx = cli_bytecode_context_alloc();
    if (!ctx) {
        cli_errmsg("Bytecode: failed to allocate bytecode context\n");
        return CL_EMEM;
    }
    rc = run_builtin_or_loaded(bcs, BC_STARTUP, builtin_bc_startup, ctx, "BC_STARTUP");
    if (rc != CL_SUCCESS) {
        cli_warnmsg("Bytecode: BC_STARTUP failed to run, disabling ALL bytecodes! Please report to https://github.com/Cisco-Talos/clamav/issues\n");
        ctx->bytecode_disable_status = 2;
    } else {
        cli_dbgmsg("Bytecode: disable status is %d\n", ctx->bytecode_disable_status);
        rc = cli_bytecode_context_getresult_int(ctx);
        /* check magic number, don't use 0 here because it is too easy for a
         * buggy bytecode to return 0 */
        if ((unsigned int)rc != (unsigned int)0xda7aba5e) {
            cli_warnmsg("Bytecode: selftest failed with code %08x. Please report to https://github.com/Cisco-Talos/clamav/issues\n",
                        rc);
            if (engine->bytecode_mode == CL_BYTECODE_MODE_TEST)
                return CL_EBYTECODE_TESTFAIL;
        }
    }
    switch (ctx->bytecode_disable_status) {
        case 1:
            if (set_mode(engine, CL_BYTECODE_MODE_INTERPRETER) == -1)
                return CL_EBYTECODE_TESTFAIL;
            break;
        case 2:
            if (set_mode(engine, CL_BYTECODE_MODE_OFF) == -1)
                return CL_EBYTECODE_TESTFAIL;
            break;
        default:
            break;
    }
    cli_bytecode_context_destroy(ctx);

    if (engine->bytecode_mode != CL_BYTECODE_MODE_INTERPRETER &&
        engine->bytecode_mode != CL_BYTECODE_MODE_OFF) {
        selfcheck(1, bcs->engine);
        rc = cli_bytecode_prepare_jit(bcs);
        if (rc == CL_SUCCESS) {
            jitok = 1;
            cli_dbgmsg("Bytecode: %u bytecode prepared with JIT\n", bcs->count);
            if (engine->bytecode_mode != CL_BYTECODE_MODE_TEST)
                return CL_SUCCESS;
        }
        if (engine->bytecode_mode == CL_BYTECODE_MODE_JIT) {
            cli_errmsg("Bytecode: JIT required, but not all bytecodes could be prepared with JIT\n");
            return CL_EMALFDB;
        }
        if (rc && engine->bytecode_mode == CL_BYTECODE_MODE_TEST) {
            cli_errmsg("Bytecode: Test mode, but not all bytecodes could be prepared with JIT\n");
            return CL_EBYTECODE_TESTFAIL;
        }
    } else {
        cli_bytecode_done_jit(bcs, 0);
    }

    if (!(dconfmask & BYTECODE_INTERPRETER)) {
        cli_dbgmsg("Bytecode: needs interpreter, but interpreter is disabled\n");
        if (set_mode(engine, CL_BYTECODE_MODE_OFF) == -1)
            return CL_EBYTECODE_TESTFAIL;
    }

    if (engine->bytecode_mode == CL_BYTECODE_MODE_OFF) {
        for (i = 0; i < bcs->count; i++)
            bcs->all_bcs[i].state = bc_disabled;
        cli_dbgmsg("Bytecode: ALL bytecodes disabled\n");
        return CL_SUCCESS;
    }

    for (i = 0; i < bcs->count; i++) {
        struct cli_bc *bc = &bcs->all_bcs[i];
        if (bc->state == bc_jit) {
            jitcount++;
            if (engine->bytecode_mode != CL_BYTECODE_MODE_TEST)
                continue;
        }
        if (bc->state == bc_interp) {
            interp++;
            continue;
        }
        rc = cli_bytecode_prepare_interpreter(bc);
        if (rc != CL_SUCCESS) {
            bc->state = bc_disabled;
            cli_warnmsg("Bytecode: %d failed to prepare for interpreter mode\n", bc->id);
            return rc;
        }
        interp++;
    }
    cli_dbgmsg("Bytecode: %u bytecode prepared with JIT, "
               "%u prepared with interpreter, %u total\n",
               jitcount, interp, bcs->count);
    return CL_SUCCESS;
}

int cli_bytecode_init(struct cli_all_bc *allbc)
{
    int ret;
    memset(allbc, 0, sizeof(*allbc));
    ret = cli_bytecode_init_jit(allbc, 0 /*XXX*/);
    cli_dbgmsg("Bytecode initialized in %s mode\n",
               allbc->engine ? "JIT" : "interpreter");
    allbc->inited = 1;
    return ret;
}

int cli_bytecode_done(struct cli_all_bc *allbc)
{
    return cli_bytecode_done_jit(allbc, 0);
}

int cli_bytecode_context_setfile(struct cli_bc_ctx *ctx, fmap_t *map)
{
    ctx->fmap           = map;
    ctx->file_size      = map->len;
    ctx->hooks.filesize = &ctx->file_size;
    return 0;
}

int cli_bytecode_runlsig(cli_ctx *cctx, struct cli_target_info *tinfo,
                         const struct cli_all_bc *bcs, unsigned bc_idx,
                         const uint32_t *lsigcnt,
                         const uint32_t *lsigsuboff, fmap_t *map)
{
    int ret;
    struct cli_bc_ctx ctx;
    const struct cli_bc *bc = &bcs->all_bcs[bc_idx - 1];
    struct cli_pe_hook_data pehookdata;

    if (bc_idx == 0)
        return CL_ENULLARG;

    memset(&ctx, 0, sizeof(ctx));
    cli_bytecode_context_setfuncid(&ctx, bc, 0);
    ctx.hooks.match_counts  = lsigcnt;
    ctx.hooks.match_offsets = lsigsuboff;
    cli_bytecode_context_setctx(&ctx, cctx);
    cli_bytecode_context_setfile(&ctx, map);
    if (tinfo && tinfo->status == 1) {
        ctx.sections = tinfo->exeinfo.sections;
        memset(&pehookdata, 0, sizeof(pehookdata));
        pehookdata.offset    = tinfo->exeinfo.offset;
        pehookdata.ep        = tinfo->exeinfo.ep;
        pehookdata.nsections = tinfo->exeinfo.nsections;
        pehookdata.hdr_size  = tinfo->exeinfo.hdr_size;
        ctx.hooks.pedata     = &pehookdata;
        ctx.resaddr          = tinfo->exeinfo.res_addr;
    }
    if (bc->hook_lsig_id) {
        cli_dbgmsg("hook lsig id %d matched (bc %d)\n", bc->hook_lsig_id, bc->id);
        /* this is a bytecode for a hook, defer running it until hook is
         * executed, so that it has all the info for the hook */
        if (cctx->hook_lsig_matches)
            cli_bitset_set(cctx->hook_lsig_matches, bc->hook_lsig_id - 1);
        /* save match counts */
        memcpy(&ctx.lsigcnt, lsigcnt, 64 * 4);
        memcpy(&ctx.lsigoff, lsigsuboff, 64 * 4);
        cli_bytecode_context_clear(&ctx);
        return CL_SUCCESS;
    }

    cli_dbgmsg("Running bytecode for logical signature match\n");
    ret = cli_bytecode_run(bcs, bc, &ctx);
    if (ret != CL_SUCCESS) {
        cli_warnmsg("Bytecode %u failed to run: %s\n", bc->id, cl_strerror(ret));
        cli_bytecode_context_clear(&ctx);
        return CL_SUCCESS;
    }
    if (ctx.virname) {
        if (cctx->num_viruses == 0) {
            int rc;
            cli_dbgmsg("Bytecode found virus: %s\n", ctx.virname);
            if (!strncmp(ctx.virname, "BC.Heuristics", 13))
                rc = cli_append_possibly_unwanted(cctx, ctx.virname);
            else
                rc = cli_append_virus(cctx, ctx.virname);
            cli_bytecode_context_clear(&ctx);
            return rc;
        } else {
            cli_bytecode_context_clear(&ctx);
            return CL_VIRUS;
        }
    }
    ret = cli_bytecode_context_getresult_int(&ctx);
    cli_dbgmsg("Bytecode %u returned code: %u\n", bc->id, ret);
    cli_bytecode_context_clear(&ctx);
    return CL_SUCCESS;
}

int cli_bytecode_runhook(cli_ctx *cctx, const struct cl_engine *engine, struct cli_bc_ctx *ctx,
                         unsigned id, fmap_t *map)
{
    const unsigned *hooks = engine->hooks[id - _BC_START_HOOKS];
    unsigned i, hooks_cnt = engine->hooks_cnt[id - _BC_START_HOOKS];
    int ret;
    unsigned executed = 0, breakflag = 0, errorflag = 0;

    if (!cctx)
        return CL_ENULLARG;

    cli_dbgmsg("Bytecode executing hook id %u (%u hooks)\n", id, hooks_cnt);
    /* restore match counts */
    cli_bytecode_context_setfile(ctx, map);
    ctx->hooks.match_counts  = ctx->lsigcnt;
    ctx->hooks.match_offsets = ctx->lsigoff;
    for (i = 0; i < hooks_cnt; i++) {
        const struct cli_bc *bc = &engine->bcs.all_bcs[hooks[i]];
        if (bc->lsig) {
            if (!cctx->hook_lsig_matches ||
                !cli_bitset_test(cctx->hook_lsig_matches, bc->hook_lsig_id - 1))
                continue;
            cli_dbgmsg("Bytecode: executing bytecode %u (lsig matched)\n", bc->id);
        }
        cli_bytecode_context_setfuncid(ctx, bc, 0);
        ret = cli_bytecode_run(&engine->bcs, bc, ctx);
        executed++;
        if (ret != CL_SUCCESS) {
            cli_warnmsg("Bytecode %u failed to run: %s\n", bc->id, cl_strerror(ret));
            errorflag = 1;
            continue;
        }
        if (ctx->virname) {
            cli_dbgmsg("Bytecode runhook found virus: %s\n", ctx->virname);
            cli_append_virus(cctx, ctx->virname);
            if (!(cctx->options->general & CL_SCAN_GENERAL_ALLMATCHES)) {
                cli_bytecode_context_clear(ctx);
                return CL_VIRUS;
            }
            cli_bytecode_context_reset(ctx);
            continue;
        }
        ret = cli_bytecode_context_getresult_int(ctx);
        /* TODO: use prefix here */
        cli_dbgmsg("Bytecode %u returned %u\n", bc->id, ret);
        if (ret == 0xcea5e) {
            cli_dbgmsg("Bytecode set BREAK flag in hook!\n");
            breakflag = 1;
        }
        if (!ret) {
            char *tempfile;
            int fd = cli_bytecode_context_getresult_file(ctx, &tempfile);
            if (fd && fd != -1) {
                if (cctx->engine->keeptmp)
                    cli_dbgmsg("Bytecode %u unpacked file saved in %s\n",
                               bc->id, tempfile);
                else
                    cli_dbgmsg("Bytecode %u unpacked file\n", bc->id);
                lseek(fd, 0, SEEK_SET);
                cli_dbgmsg("***** Scanning unpacked file ******\n");

                ret = cli_magic_scan_desc(fd, tempfile, cctx, NULL);

                if (!cctx->engine->keeptmp)
                    if (ftruncate(fd, 0) == -1)
                        cli_dbgmsg("ftruncate failed on %d\n", fd);
                close(fd);
                if (!cctx->engine->keeptmp) {
                    if (tempfile && cli_unlink(tempfile))
                        ret = CL_EUNLINK;
                }
                free(tempfile);
                if (ret != CL_CLEAN) {
                    if (ret == CL_VIRUS) {
                        cli_dbgmsg("Scanning unpacked file by bytecode %u found a virus\n", bc->id);
                        if (cctx->options->general & CL_SCAN_GENERAL_ALLMATCHES) {
                            cli_bytecode_context_reset(ctx);
                            continue;
                        }
                        cli_bytecode_context_clear(ctx);
                        return ret;
                    }
                }
                cli_bytecode_context_reset(ctx);
                continue;
            }
        }
        cli_bytecode_context_reset(ctx);
    }
    if (executed)
        cli_dbgmsg("Bytecode: executed %u bytecodes for this hook\n", executed);
    else
        cli_dbgmsg("Bytecode: no logical signature matched, no bytecode executed\n");
    if (errorflag && cctx->engine->bytecode_mode == CL_BYTECODE_MODE_TEST)
        return CL_EBYTECODE_TESTFAIL;
    return breakflag ? CL_BREAK : CL_CLEAN;
}

int cli_bytecode_context_setpe(struct cli_bc_ctx *ctx, const struct cli_pe_hook_data *data, const struct cli_exe_section *sections)
{
    ctx->sections     = sections;
    ctx->hooks.pedata = data;
    return 0;
}

void cli_bytecode_context_setctx(struct cli_bc_ctx *ctx, void *cctx)
{
    ctx->ctx              = cctx;
    ctx->bytecode_timeout = ((cli_ctx *)cctx)->engine->bytecode_timeout;
}

void cli_bytecode_describe(const struct cli_bc *bc)
{
    char buf[128];
    int cols;
    unsigned i;
    time_t stamp;
    int had;

    if (!bc) {
        printf("(null bytecode)\n");
        return;
    }

    stamp = bc->metadata.timestamp;
    printf("Bytecode format functionality level: %u\n", bc->metadata.formatlevel);
    printf("Bytecode metadata:\n\tcompiler version: %s\n",
           bc->metadata.compiler ? bc->metadata.compiler : "N/A");
    printf("\tcompiled on: (%d) %s",
           (uint32_t)stamp,
           cli_ctime(&stamp, buf, sizeof(buf)));
    printf("\tcompiled by: %s\n", bc->metadata.sigmaker ? bc->metadata.sigmaker : "N/A");
    /*TODO: parse and display arch name, also take it into account when
      JITing*/
    printf("\ttarget exclude: %d\n", bc->metadata.targetExclude);
    printf("\tbytecode type: ");
    switch (bc->kind) {
        case BC_GENERIC:
            puts("generic, not loadable by clamscan/clamd");
            break;
        case BC_STARTUP:
            puts("run on startup (unique)");
            break;
        case BC_LOGICAL:
            puts("logical only");
            break;
        case BC_PE_UNPACKER:
            puts("PE unpacker hook");
            break;
        case BC_PE_ALL:
            puts("all PE hook");
            break;
        case BC_PRECLASS:
            puts("preclass hook");
            break;
        case BC_ELF_UNPACKER:
            puts("ELF unpacker hook");
            break;
        case BC_MACHO_UNPACKER:
            puts("Mach-O unpacker hook");
            break;
        default:
            printf("Unknown (type %u)", bc->kind);
            break;
    }
    /* 0 means no limit */
    printf("\tbytecode functionality level: %u - %u\n",
           bc->metadata.minfunc, bc->metadata.maxfunc);
    printf("\tbytecode logical signature: %s\n",
           bc->lsig ? bc->lsig : "<none>");
    printf("\tvirusname prefix: %s\n",
           bc->vnameprefix);
    printf("\tvirusnames: %u\n", bc->vnames_cnt);
    printf("\tbytecode triggered on: ");
    switch (bc->kind) {
        case BC_GENERIC:
            puts("N/A (loaded in clambc only)");
            break;
        case BC_LOGICAL:
            puts("files matching logical signature");
            break;
        case BC_PE_UNPACKER:
            if (bc->lsig)
                puts("PE files matching logical signature (unpacked)");
            else
                puts("all PE files! (unpacked)");
            break;
        case BC_PDF:
            puts("PDF files");
            break;
        case BC_PE_ALL:
            if (bc->lsig)
                puts("PE files matching logical signature");
            else
                puts("all PE files!");
            break;
        case BC_PRECLASS:
            if (bc->lsig)
                puts("PRECLASS files matching logical signature");
            else
                puts("all PRECLASS files!");
            break;
        case BC_ELF_UNPACKER:
            if (bc->lsig)
                puts("ELF files matching logical signature (unpacked)");
            else
                puts("all ELF files! (unpacked)");
            break;
        case BC_MACHO_UNPACKER:
            if (bc->lsig)
                puts("Mach-O files matching logical signature (unpacked)");
            else
                puts("all Mach-O files! (unpacked)");
            break;
        default:
            puts("N/A (unknown type)\n");
            break;
    }
    printf("\tnumber of functions: %u\n\tnumber of types: %u\n",
           bc->num_func, bc->num_types);
    printf("\tnumber of global constants: %u\n", (unsigned)bc->num_globals);
    printf("\tnumber of debug nodes: %u\n", bc->dbgnode_cnt);
    printf("\tbytecode APIs used:");
    cols = 0; /* remaining */
    had  = 0;
    for (i = 0; i < cli_apicall_maxapi; i++) {
        if (cli_bitset_test(bc->uses_apis, i)) {
            unsigned len = strlen(cli_apicalls[i].name);
            if (had)
                printf(",");
            if (len > (unsigned int)cols) {
                printf("\n\t");
                cols = 72;
            }
            printf(" %s", cli_apicalls[i].name);
            had = 1;
            cols -= len;
        }
    }
    printf("\n");
}

const char *bc_tystr[] = {
    "DFunctionType",
    "DPointerType",
    "DStructType",
    "DPackedStructType",
    "DArrayType"};

const char *bc_opstr[] = {
    "OP_BC_NULL",
    "OP_BC_ADD", /* =1*/
    "OP_BC_SUB",
    "OP_BC_MUL",
    "OP_BC_UDIV",
    "OP_BC_SDIV",
    "OP_BC_UREM",
    "OP_BC_SREM",
    "OP_BC_SHL",
    "OP_BC_LSHR",
    "OP_BC_ASHR",
    "OP_BC_AND",
    "OP_BC_OR",
    "OP_BC_XOR",

    "OP_BC_TRUNC",
    "OP_BC_SEXT",
    "OP_BC_ZEXT",

    "OP_BC_BRANCH",
    "OP_BC_JMP",
    "OP_BC_RET",
    "OP_BC_RET_VOID",

    "OP_BC_ICMP_EQ",
    "OP_BC_ICMP_NE",
    "OP_BC_ICMP_UGT",
    "OP_BC_ICMP_UGE",
    "OP_BC_ICMP_ULT",
    "OP_BC_ICMP_ULE",
    "OP_BC_ICMP_SGT",
    "OP_BC_ICMP_SGE",
    "OP_BC_ICMP_SLE",
    "OP_BC_ICMP_SLT",
    "OP_BC_SELECT",
    "OP_BC_CALL_DIRECT",
    "OP_BC_CALL_API",
    "OP_BC_COPY",
    "OP_BC_GEP1",
    "OP_BC_GEPZ",
    "OP_BC_GEPN",
    "OP_BC_STORE",
    "OP_BC_LOAD",
    "OP_BC_MEMSET",
    "OP_BC_MEMCPY",
    "OP_BC_MEMMOVE",
    "OP_BC_MEMCMP",
    "OP_BC_ISBIGENDIAN",
    "OP_BC_ABORT",
    "OP_BC_BSWAP16",
    "OP_BC_BSWAP32",
    "OP_BC_BSWAP64",
    "OP_BC_PTRDIFF32",
    "OP_BC_PTRTOINT64",
    "OP_BC_INVALID" /* last */
};

extern unsigned cli_numapicalls;
static void cli_bytetype_helper(const struct cli_bc *bc, unsigned tid)
{
    unsigned i, j;
    const struct cli_bc_type *ty;

    if (tid & 0x8000) {
        printf("alloc ");
        tid &= 0x7fff;
    }

    if (tid < 65) {
        printf("i%d", tid);
        return;
    }

    i = tid - 65;
    if (i >= bc->num_types) {
        printf("invalid type");
        return;
    }
    ty = &bc->types[i];

    switch (ty->kind) {
        case DFunctionType:
            cli_bytetype_helper(bc, ty->containedTypes[0]);
            printf(" func ( ");
            for (j = 1; j < ty->numElements; ++j) {
                cli_bytetype_helper(bc, ty->containedTypes[0]);
                printf(" ");
            }
            printf(")");
            break;
        case DPointerType:
            cli_bytetype_helper(bc, ty->containedTypes[0]);
            printf("*");
            break;
        case DStructType:
        case DPackedStructType:
            printf("{ ");
            for (j = 0; j < ty->numElements; ++j) {
                cli_bytetype_helper(bc, ty->containedTypes[0]);
                printf(" ");
            }
            printf("}");
            break;
        case DArrayType:
            printf("[");
            printf("%d x ", ty->numElements);
            cli_bytetype_helper(bc, ty->containedTypes[0]);
            printf("]");
            break;
        default:
            printf("unhandled type kind %d, cannot parse", ty->kind);
            break;
    }
}

void cli_bytetype_describe(const struct cli_bc *bc)
{
    unsigned i, tid;

    printf("found %d extra types of %d total, starting at tid %d\n",
           bc->num_types, 64 + bc->num_types, bc->start_tid);

    printf("TID  KIND                INTERNAL\n");
    printf("------------------------------------------------------------------------\n");
    for (i = 0, tid = 65; i < bc->num_types - 1; ++i, ++tid) {
        printf("%3d: %-20s", tid, bc_tystr[bc->types[i].kind]);
        cli_bytetype_helper(bc, tid);
        printf("\n");
    }
    printf("------------------------------------------------------------------------\n");
}

void cli_bytevalue_describe(const struct cli_bc *bc, unsigned funcid)
{
    unsigned i, total = 0;
    const struct cli_bc_func *func;

    if (funcid >= bc->num_func) {
        printf("bytecode diagnostic: funcid [%u] outside bytecode numfuncs [%u]\n",
               funcid, bc->num_func);
        return;
    }
    // globals
    printf("found a total of %zu globals\n", bc->num_globals);
    printf("GID  ID    VALUE\n");
    printf("------------------------------------------------------------------------\n");
    for (i = 0; i < bc->num_globals; ++i) {
        printf("%3u [%3u]: ", i, i);
        cli_bytetype_helper(bc, bc->globaltys[i]);
        printf(" unknown\n");
    }
    printf("------------------------------------------------------------------------\n");

    // arguments and local values
    func = &bc->funcs[funcid];
    printf("found %d values with %d arguments and %d locals\n",
           func->numValues, func->numArgs, func->numLocals);
    printf("VID  ID    VALUE\n");
    printf("------------------------------------------------------------------------\n");
    for (i = 0; i < func->numValues; ++i) {
        printf("%3u [%3u]: ", i, total++);
        cli_bytetype_helper(bc, func->types[i]);
        if (i < func->numArgs)
            printf(" argument");
        printf("\n");
    }
    printf("------------------------------------------------------------------------\n");

    // constants
    printf("found a total of %d constants\n", func->numConstants);
    printf("CID  ID    VALUE\n");
    printf("------------------------------------------------------------------------\n");
    for (i = 0; i < func->numConstants; ++i) {
        printf("%3u [%3u]: " STDu64 "(0x" STDx64 ")\n", i, total++, func->constants[i], func->constants[i]);
    }
    printf("------------------------------------------------------------------------\n");
    printf("found a total of %u total values\n", total);
    printf("------------------------------------------------------------------------\n");
    return;
}

void cli_byteinst_describe(const struct cli_bc_inst *inst, unsigned *bbnum)
{
    size_t j;
    char inst_str[256];
    const struct cli_apicall *api;

    if (inst->opcode > OP_BC_INVALID) {
        printf("opcode %u[%u] of type %u is not implemented yet!",
               inst->opcode, inst->interp_op / 5, inst->interp_op % 5);
        return;
    }

    snprintf(inst_str, sizeof(inst_str), "%-20s[%-3d/%3d/%3d]", bc_opstr[inst->opcode],
             inst->opcode, inst->interp_op, inst->interp_op % inst->opcode);
    printf("%-35s", inst_str);
    switch (inst->opcode) {
            // binary operations
        case OP_BC_ADD:
            printf("%d = %d + %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_SUB:
            printf("%d = %d - %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_MUL:
            printf("%d = %d * %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_UDIV:
            printf("%d = %d / %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_SDIV:
            printf("%d = %d / %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_UREM:
            printf("%d = %d %% %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_SREM:
            printf("%d = %d %% %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_SHL:
            printf("%d = %d << %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_LSHR:
            printf("%d = %d >> %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ASHR:
            printf("%d = %d >> %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_AND:
            printf("%d = %d & %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_OR:
            printf("%d = %d | %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_XOR:
            printf("%d = %d ^ %d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;

            // casting operations
        case OP_BC_TRUNC:
            printf("%d = %d trunc " STDx64, inst->dest, inst->u.cast.source, inst->u.cast.mask);
            break;
        case OP_BC_SEXT:
            printf("%d = %d sext " STDx64, inst->dest, inst->u.cast.source, inst->u.cast.mask);
            break;
        case OP_BC_ZEXT:
            printf("%d = %d zext " STDx64, inst->dest, inst->u.cast.source, inst->u.cast.mask);
            break;

            // control operations (termination instructions)
        case OP_BC_BRANCH:
            printf("br %d ? bb.%d : bb.%d", inst->u.branch.condition,
                   inst->u.branch.br_true, inst->u.branch.br_false);
            (*bbnum)++;
            break;
        case OP_BC_JMP:
            printf("jmp bb.%d", inst->u.jump);
            (*bbnum)++;
            break;
        case OP_BC_RET:
            printf("ret %d", inst->u.unaryop);
            (*bbnum)++;
            break;
        case OP_BC_RET_VOID:
            printf("ret void");
            (*bbnum)++;
            break;

            // comparison operations
        case OP_BC_ICMP_EQ:
            printf("%d = (%d == %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_NE:
            printf("%d = (%d != %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_UGT:
            printf("%d = (%d > %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_UGE:
            printf("%d = (%d >= %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_ULT:
            printf("%d = (%d < %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_ULE:
            printf("%d = (%d >= %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_SGT:
            printf("%d = (%d > %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_SGE:
            printf("%d = (%d >= %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_SLE:
            printf("%d = (%d <= %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_ICMP_SLT:
            printf("%d = (%d < %d)", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_SELECT:
            printf("%d = %d ? %d : %d)", inst->dest, inst->u.three[0],
                   inst->u.three[1], inst->u.three[2]);
            break;

            // function calling
        case OP_BC_CALL_DIRECT:
            printf("%d = call F.%d (", inst->dest, inst->u.ops.funcid);
            for (j = 0; j < inst->u.ops.numOps; ++j) {
                if (j == inst->u.ops.numOps - 1) {
                    printf("%d", inst->u.ops.ops[j]);
                } else {
                    printf("%d, ", inst->u.ops.ops[j]);
                }
            }
            printf(")");
            break;
        case OP_BC_CALL_API: {
            if (inst->u.ops.funcid > cli_numapicalls) {
                printf("apicall FID %d not yet implemented!\n", inst->u.ops.funcid);
                break;
            }
            api = &cli_apicalls[inst->u.ops.funcid];
            switch (api->kind) {
                case 0:
                    printf("%d = %s[%d] (%d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1]);
                    break;
                case 1:
                    printf("%d = %s[%d] (p.%d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1]);
                    break;
                case 2:
                    printf("%d = %s[%d] (%d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0]);
                    break;
                case 3:
                    printf("p.%d = %s[%d] (%d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0]);
                    break;
                case 4:
                    printf("%d = %s[%d] (p.%d, %d, %d, %d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1],
                           inst->u.ops.ops[2], inst->u.ops.ops[3], inst->u.ops.ops[4]);
                    break;
                case 5:
                    printf("%d = %s[%d] ()", inst->dest, api->name,
                           inst->u.ops.funcid);
                    break;
                case 6:
                    printf("p.%d = %s[%d] (%d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1]);
                    break;
                case 7:
                    printf("%d = %s[%d] (%d, %d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1],
                           inst->u.ops.ops[2]);
                    break;
                case 8:
                    printf("%d = %s[%d] (p.%d, %d, p.%d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1],
                           inst->u.ops.ops[2], inst->u.ops.ops[3]);
                    break;
                case 9:
                    printf("%d = %s[%d] (p.%d, %d, %d)", inst->dest, api->name,
                           inst->u.ops.funcid, inst->u.ops.ops[0], inst->u.ops.ops[1],
                           inst->u.ops.ops[2]);
                    break;
                default:
                    printf("type %u apicalls not yet implemented!\n", api->kind);
                    break;
            }
        } break;

            // memory operations
        case OP_BC_COPY:
            printf("cp %d -> %d", inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_GEP1:
            printf("%d = gep1 p.%d + (%d * %d)", inst->dest, inst->u.three[1],
                   inst->u.three[2], inst->u.three[0]);
            break;
        case OP_BC_GEPZ:
            printf("%d = gepz p.%d + (%d)", inst->dest,
                   inst->u.three[1], inst->u.three[2]);
            break;
        case OP_BC_GEPN:
            printf("illegal opcode, impossible");
            break;
        case OP_BC_STORE:
            printf("store %d -> p.%d", inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_LOAD:
            printf("load  %d <- p.%d", inst->dest, inst->u.unaryop);
            break;

            // llvm intrinsics
        case OP_BC_MEMSET:
            printf("%d = memset (p.%d, %d, %d)", inst->dest, inst->u.three[0],
                   inst->u.three[1], inst->u.three[2]);
            break;
        case OP_BC_MEMCPY:
            printf("%d = memcpy (p.%d, p.%d, %d)", inst->dest, inst->u.three[0],
                   inst->u.three[1], inst->u.three[2]);
            break;
        case OP_BC_MEMMOVE:
            printf("%d = memmove (p.%d, p.%d, %d)", inst->dest, inst->u.three[0],
                   inst->u.three[1], inst->u.three[2]);
            break;
        case OP_BC_MEMCMP:
            printf("%d = memcmp (p.%d, p.%d, %d)", inst->dest, inst->u.three[0],
                   inst->u.three[1], inst->u.three[2]);
            break;

            // utility operations
        case OP_BC_ISBIGENDIAN:
            printf("%d = isbigendian()", inst->dest);
            break;
        case OP_BC_ABORT:
            printf("ABORT!!");
            break;
        case OP_BC_BSWAP16:
            printf("%d = bswap16 %d", inst->dest, inst->u.unaryop);
            break;
        case OP_BC_BSWAP32:
            printf("%d = bswap32 %d", inst->dest, inst->u.unaryop);
            break;
        case OP_BC_BSWAP64:
            printf("%d = bswap64 %d", inst->dest, inst->u.unaryop);
            break;
        case OP_BC_PTRDIFF32:
            printf("%d = ptrdiff32 p.%d p.%d", inst->dest, inst->u.binop[0], inst->u.binop[1]);
            break;
        case OP_BC_PTRTOINT64:
            printf("%d = ptrtoint64 p.%d", inst->dest, inst->u.unaryop);
            break;
        case OP_BC_INVALID: /* last */
            printf("INVALID!!");
            break;

        default:
            // redundant check
            printf("opcode %u[%u] of type %u is not implemented yet!",
                   inst->opcode, inst->interp_op / 5, inst->interp_op % 5);
            break;
    }
}

void cli_bytefunc_describe(const struct cli_bc *bc, unsigned funcid)
{
    unsigned i, bbnum, bbpre;
    const struct cli_bc_func *func;

    if (funcid >= bc->num_func) {
        printf("bytecode diagnostic: funcid [%u] outside bytecode numfuncs [%u]\n",
               funcid, bc->num_func);
        return;
    }

    func = &bc->funcs[funcid];

    printf("FUNCTION ID: F.%d -> NUMINSTS %d\n", funcid, func->numInsts);
    printf("BB   IDX  OPCODE              [ID /IID/MOD]  INST\n");
    printf("------------------------------------------------------------------------\n");
    bbpre = 0;
    bbnum = 0;
    for (i = 0; i < func->numInsts; ++i) {
        if (bbpre != bbnum) {
            printf("\n");
            bbpre = bbnum;
        }

        printf("%3d  %3d  ", bbnum, i);
        cli_byteinst_describe(&func->allinsts[i], &bbnum);
        printf("\n");
    }
    printf("------------------------------------------------------------------------\n");
}