1
mirror of https://github.com/hashcat/hashcat synced 2024-11-10 02:34:43 +01:00

Encoding: Truncate password candidates in UTF8 -> UTF16 conversion if it contains an invalid UTF8 byte sequence

This commit is contained in:
Jens Steube 2021-08-11 16:44:04 +02:00
parent 577fada1de
commit f72f2b0f09
2 changed files with 174 additions and 0 deletions

View File

@ -1711,6 +1711,15 @@ DECLSPEC u32 hc_bfe_S (const u32 a, const u32 b, const u32 c)
return r;
}
DECLSPEC u32 hc_bytealign_be_S (const u32 a, const u32 b, const int c)
{
const int c_mod_4 = c & 3;
const u32 r = hc_byte_perm_S (b, a, (0x76543210 >> (c_mod_4 * 4)) & 0xffff);
return r;
}
DECLSPEC u32x hc_bytealign (const u32x a, const u32x b, const int c)
{
const int c_mod_4 = c & 3;
@ -2056,6 +2065,152 @@ DECLSPEC int hc_enc_has_next (hc_enc_t *hc_enc, const int sz)
return 0;
}
DECLSPEC int hc_enc_validate_utf8 (const u32 *src_buf, const int src_pos, const int extraBytesToRead)
{
const u8 *src_ptr = (const u8 *) src_buf;
if (extraBytesToRead == 0)
{
const u8 c0 = src_ptr[src_pos + 0]; if (c0 >= 0x80) return 0;
}
else if (extraBytesToRead == 1)
{
const u8 c0 = src_ptr[src_pos + 0]; if ((c0 < 0xc2) || (c0 > 0xdf)) return 0;
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
}
else if (extraBytesToRead == 2)
{
const u8 c0 = src_ptr[src_pos + 0];
if ((c0 >= 0xe0) && (c0 <= 0xe0))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0xa0) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else if ((c0 >= 0xe1) && (c0 <= 0xec))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else if ((c0 >= 0xed) && (c0 <= 0xed))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0x9f)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else if ((c0 >= 0xee) && (c0 <= 0xef))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else
{
return 0;
}
}
else if (extraBytesToRead == 3)
{
const u8 c0 = src_ptr[src_pos + 0];
if ((c0 >= 0xf0) && (c0 <= 0xf0))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x90) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
const u8 c3 = src_ptr[src_pos + 3]; if ((c3 < 0x80) || (c3 > 0xbf)) return 0;
}
else if ((c0 >= 0xf1) && (c0 <= 0xf3))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
const u8 c3 = src_ptr[src_pos + 3]; if ((c3 < 0x80) || (c3 > 0xbf)) return 0;
}
else if ((c0 >= 0xf4) && (c0 <= 0xf4))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
const u8 c3 = src_ptr[src_pos + 3]; if ((c3 < 0x80) || (c3 > 0xbf)) return 0;
}
else
{
return 0;
}
}
return 1;
}
DECLSPEC int hc_enc_validate_utf8_global (GLOBAL_AS const u32 *src_buf, const int src_pos, const int extraBytesToRead)
{
GLOBAL_AS const u8 *src_ptr = (GLOBAL_AS const u8 *) src_buf;
if (extraBytesToRead == 0)
{
const u8 c0 = src_ptr[src_pos + 0]; if (c0 >= 0x80) return 0;
}
else if (extraBytesToRead == 1)
{
const u8 c0 = src_ptr[src_pos + 0]; if ((c0 < 0xc2) || (c0 > 0xdf)) return 0;
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
}
else if (extraBytesToRead == 2)
{
const u8 c0 = src_ptr[src_pos + 0];
if ((c0 >= 0xe0) && (c0 <= 0xe0))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0xa0) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else if ((c0 >= 0xe1) && (c0 <= 0xec))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else if ((c0 >= 0xed) && (c0 <= 0xed))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0x9f)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else if ((c0 >= 0xee) && (c0 <= 0xef))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
}
else
{
return 0;
}
}
else if (extraBytesToRead == 3)
{
const u8 c0 = src_ptr[src_pos + 0];
if ((c0 >= 0xf0) && (c0 <= 0xf0))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x90) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
const u8 c3 = src_ptr[src_pos + 3]; if ((c3 < 0x80) || (c3 > 0xbf)) return 0;
}
else if ((c0 >= 0xf1) && (c0 <= 0xf3))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
const u8 c3 = src_ptr[src_pos + 3]; if ((c3 < 0x80) || (c3 > 0xbf)) return 0;
}
else if ((c0 >= 0xf4) && (c0 <= 0xf4))
{
const u8 c1 = src_ptr[src_pos + 1]; if ((c1 < 0x80) || (c1 > 0xbf)) return 0;
const u8 c2 = src_ptr[src_pos + 2]; if ((c2 < 0x80) || (c2 > 0xbf)) return 0;
const u8 c3 = src_ptr[src_pos + 3]; if ((c3 < 0x80) || (c3 > 0xbf)) return 0;
}
else
{
return 0;
}
}
return 1;
}
// Input buffer and Output buffer size has to be multiple of 4 and at least of size 4.
// The output buffer is not zero padded, so entire buffer has to be set all zero before entering this function or truncated afterwards.
@ -2115,6 +2270,15 @@ DECLSPEC int hc_enc_next (hc_enc_t *hc_enc, const u32 *src_buf, const int src_le
return dst_pos;
}
if (hc_enc_validate_utf8 (src_buf, src_pos, extraBytesToRead) == 0)
{
// broken input
hc_enc->pos = src_len;
return dst_pos;
}
u32 ch = 0;
switch (extraBytesToRead)
@ -2252,6 +2416,15 @@ DECLSPEC int hc_enc_next_global (hc_enc_t *hc_enc, GLOBAL_AS const u32 *src_buf,
return dst_pos;
}
if (hc_enc_validate_utf8_global (src_buf, src_pos, extraBytesToRead) == 0)
{
// broken input
hc_enc->pos = src_len;
return dst_pos;
}
u32 ch = 0;
switch (extraBytesToRead)

View File

@ -55,6 +55,7 @@
- Constants: Make const char * pointers actually const char * const pointers
- Deprecated Plugins: Add new module function module_deprecated_notice() to mark a plugin as deprecated and to return a free text user notice
- Deprecated Plugins: Marked plugins 2500/2501 and 16800/16801 as deprecated
- Encoding: Truncate password candidates in UTF8 -> UTF16 conversion if it contains an invalid UTF8 byte sequence
- Filehandling: Use const char for fopen mode to fix -Wwrite-strings warnings
- Hardware Monitor: Added support for OverDrive 7 and 8 based GPUs
- HIP Kernels: Dependency on hip/hip runtime.h has been removed to enable easier integration of the HIP backend under Windows