[PATCH v9 0/6] lib and lib/cmdline enhancements

[PATCH v9 0/6] lib and lib/cmdline enhancements

[Dmitry Antipov]

Adjust '_parse_integer_limit()' and 'memparse()' to not ignore overflows,
extend string to 64-bit integer conversion tests, add KUnit-based test
for 'memparse()', fix kernel-doc glitches found in lib/cmdline.c and
adjust riscv32 linker script to export symbols needed for EFI stub.

Dmitry Antipov (6):
  lib: fix _parse_integer_limit() to handle overflow
  lib: fix memparse() to handle overflow
  lib: add more string to 64-bit integer conversion overflow tests
  lib/cmdline_kunit: add test case for memparse()
  lib/cmdline: adjust a few comments to fix kernel-doc -Wreturn warnings
  riscv: export symbols needed for riscv32 EFI stub

 arch/riscv/kernel/image-vars.h |  9 ++++++
 lib/cmdline.c                  | 30 ++++++++++++------
 lib/kstrtox.c                  | 37 +++++++++++++---------
 lib/test-kstrtox.c             |  6 ++++
 lib/tests/cmdline_kunit.c      | 56 ++++++++++++++++++++++++++++++++++
 5 files changed, 113 insertions(+), 25 deletions(-)

-- 
2.53.0

_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

[PATCH v9 1/6] lib: fix _parse_integer_limit() to handle overflow

[Dmitry Antipov]

In '_parse_integer_limit()', adjust native integer arithmetic
with near-to-overflow branch where 'check_mul_overflow()' and
'check_add_overflow()' are used to check whether an intermediate
result goes out of range, and denote such a case with ULLONG_MAX,
thus making the function more similar to standard C library's
'strtoull()'. Adjust comment to kernel-doc style as well.

Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
---
v9: bump version to match the series
v8: do not use explicit temporary in check_xxx_overflow()
    calls and handle overflow flag using separate variable
v7: drop redundant check against ULLONG_MAX and restore
    original comment
v6: more compact for-loop and minor style adjustments again
v5: minor brace style adjustment
v4: restore plain integer arithmetic and use check_xxx_overflow()
    on near-to-overflow branch only
v3: adjust commit message and comments as suggested by Andy
v2: initial version to join the series
---
 lib/kstrtox.c | 37 ++++++++++++++++++++++---------------
 1 file changed, 22 insertions(+), 15 deletions(-)

diff --git a/lib/kstrtox.c b/lib/kstrtox.c
index 97be2a39f537..edc4eb7c1bca 100644
--- a/lib/kstrtox.c
+++ b/lib/kstrtox.c
@@ -39,25 +39,30 @@ const char *_parse_integer_fixup_radix(const char *s, unsigned int *base)
 	return s;
 }
 
-/*
- * Convert non-negative integer string representation in explicitly given radix
- * to an integer. A maximum of max_chars characters will be converted.
+/**
+ * _parse_integer_limit - Convert integer string representation to an integer
+ * @s: Integer string representation
+ * @base: Radix
+ * @p: Where to store result
+ * @max_chars: Maximum amount of characters to convert
+ *
+ * Convert non-negative integer string representation in explicitly given
+ * radix to an integer. If overflow occurs, value at @p is set to ULLONG_MAX.
  *
- * Return number of characters consumed maybe or-ed with overflow bit.
- * If overflow occurs, result integer (incorrect) is still returned.
+ * This function is the workhorse of other string conversion functions and it
+ * is discouraged to use it explicitly. Consider kstrto*() family instead.
  *
- * Don't you dare use this function.
+ * Return: Number of characters consumed, maybe ORed with overflow bit
  */
 noinline
 unsigned int _parse_integer_limit(const char *s, unsigned int base, unsigned long long *p,
 				  size_t max_chars)
 {
+	unsigned int rv, overflow = 0;
 	unsigned long long res;
-	unsigned int rv;
 
 	res = 0;
-	rv = 0;
-	while (max_chars--) {
+	for (rv = 0; rv < max_chars; rv++, s++) {
 		unsigned int c = *s;
 		unsigned int lc = _tolower(c);
 		unsigned int val;
@@ -76,15 +81,17 @@ unsigned int _parse_integer_limit(const char *s, unsigned int base, unsigned lon
 		 * it in the max base we support (16)
 		 */
 		if (unlikely(res & (~0ull << 60))) {
-			if (res > div_u64(ULLONG_MAX - val, base))
-				rv |= KSTRTOX_OVERFLOW;
+			if (check_mul_overflow(res, base, &res) ||
+			    check_add_overflow(res, val, &res)) {
+				res = ULLONG_MAX;
+				overflow = KSTRTOX_OVERFLOW;
+			}
+		} else {
+			res = res * base + val;
 		}
-		res = res * base + val;
-		rv++;
-		s++;
 	}
 	*p = res;
-	return rv;
+	return rv | overflow;
 }
 
 noinline
-- 
2.53.0


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

[PATCH v9 2/6] lib: fix memparse() to handle overflow

[Dmitry Antipov]

Since '_parse_integer_limit()' (and so 'simple_strtoull()') is now
capable to handle overflow, adjust 'memparse()' to handle overflow
(denoted by ULLONG_MAX) returned from 'simple_strtoull()'. Also
use 'check_shl_overflow()' to catch an overflow possibly caused
by processing size suffix and denote it with ULLONG_MAX as well.

Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
---
v9: bump version to match the series
v8: do not use temporary in check_shl_overflow()
v7: do not double-adjust endptr and drop
    redundant check against ULLONG_MAX
v6: handle valid-suffix-only string like "k"
    as unrecognized, minor style adjustments
v5: initial version to join the series
---
 lib/cmdline.c | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)

diff --git a/lib/cmdline.c b/lib/cmdline.c
index 90ed997d9570..f6e4b113ca9f 100644
--- a/lib/cmdline.c
+++ b/lib/cmdline.c
@@ -150,39 +150,46 @@ EXPORT_SYMBOL(get_options);
 unsigned long long memparse(const char *ptr, char **retptr)
 {
 	char *endptr;	/* local pointer to end of parsed string */
-
 	unsigned long long ret = simple_strtoull(ptr, &endptr, 0);
+	unsigned int shl = 0;
 
+	/* Consume valid suffix even in case of overflow. */
 	switch (*endptr) {
 	case 'E':
 	case 'e':
-		ret <<= 10;
+		shl += 10;
 		fallthrough;
 	case 'P':
 	case 'p':
-		ret <<= 10;
+		shl += 10;
 		fallthrough;
 	case 'T':
 	case 't':
-		ret <<= 10;
+		shl += 10;
 		fallthrough;
 	case 'G':
 	case 'g':
-		ret <<= 10;
+		shl += 10;
 		fallthrough;
 	case 'M':
 	case 'm':
-		ret <<= 10;
+		shl += 10;
 		fallthrough;
 	case 'K':
 	case 'k':
-		ret <<= 10;
-		endptr++;
+		shl += 10;
 		fallthrough;
 	default:
 		break;
 	}
 
+	if (shl && likely(ptr != endptr)) {
+		/* Have valid suffix with preceding number. */
+		if (unlikely(check_shl_overflow(ret, shl, &ret)))
+			ret = ULLONG_MAX;
+		endptr++;
+	}
+
 	if (retptr)
 		*retptr = endptr;
 
-- 
2.53.0


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

[PATCH v9 3/6] lib: add more string to 64-bit integer conversion overflow tests

[Dmitry Antipov]

Add a few more string to 64-bit integer conversion tests to
check whether 'kstrtoull()', 'kstrtoll()', 'kstrtou64()' and
'kstrtos64()' can handle overflows reported by
'_parse_integer_limit()'.

Suggested-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
---
v5 and upwards: bump version to match the series
v4: initial version to join the series
---
 lib/test-kstrtox.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/lib/test-kstrtox.c b/lib/test-kstrtox.c
index ee87fef66cb5..811128d0df16 100644
--- a/lib/test-kstrtox.c
+++ b/lib/test-kstrtox.c
@@ -198,6 +198,7 @@ static void __init test_kstrtoull_fail(void)
 		{"10000000000000000000000000000000000000000000000000000000000000000",	2},
 		{"2000000000000000000000",	8},
 		{"18446744073709551616",	10},
+		{"569202370375329612767",	10},
 		{"10000000000000000",	16},
 		/* negative */
 		{"-0", 0},
@@ -275,9 +276,11 @@ static void __init test_kstrtoll_fail(void)
 		{"9223372036854775809",	10},
 		{"18446744073709551614",	10},
 		{"18446744073709551615",	10},
+		{"569202370375329612767",	10},
 		{"-9223372036854775809",	10},
 		{"-18446744073709551614",	10},
 		{"-18446744073709551615",	10},
+		{"-569202370375329612767",	10},
 		/* sign is first character if any */
 		{"-+1", 0},
 		{"-+1", 8},
@@ -334,6 +337,7 @@ static void __init test_kstrtou64_fail(void)
 		{"-1",	10},
 		{"18446744073709551616",	10},
 		{"18446744073709551617",	10},
+		{"569202370375329612767",	10},
 	};
 	TEST_FAIL(kstrtou64, u64, "%llu", test_u64_fail);
 }
@@ -386,6 +390,8 @@ static void __init test_kstrtos64_fail(void)
 		{"18446744073709551615",	10},
 		{"18446744073709551616",	10},
 		{"18446744073709551617",	10},
+		{"569202370375329612767",	10},
+		{"-569202370375329612767",	10},
 	};
 	TEST_FAIL(kstrtos64, s64, "%lld", test_s64_fail);
 }
-- 
2.53.0


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

[PATCH v9 4/6] lib/cmdline_kunit: add test case for memparse()

[Dmitry Antipov]

Better late than never, now there is a long-awaited basic
test for 'memparse()' which is provided by cmdline.c.

Suggested-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
---
v7 and upwards: bump version to match the series
v6: tests to check whether valid-suffix-only string is handled as unrecognized
v5: even more tests to trigger overflow with size suffix
v4: move actual overflow tests to test-kstrtox.c
v3: adjust style as suggested by Andy
v2: few more test cases to trigger overflows
---
 lib/tests/cmdline_kunit.c | 56 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 56 insertions(+)

diff --git a/lib/tests/cmdline_kunit.c b/lib/tests/cmdline_kunit.c
index c1602f797637..18b19c85baef 100644
--- a/lib/tests/cmdline_kunit.c
+++ b/lib/tests/cmdline_kunit.c
@@ -6,6 +6,7 @@
 #include <kunit/test.h>
 #include <linux/kernel.h>
 #include <linux/random.h>
+#include <linux/sizes.h>
 #include <linux/string.h>
 
 static const char *cmdline_test_strings[] = {
@@ -139,11 +140,66 @@ static void cmdline_test_range(struct kunit *test)
 	} while (++i < ARRAY_SIZE(cmdline_test_range_strings));
 }
 
+struct cmdline_test_memparse_entry {
+	const char *input;
+	const char *unrecognized;
+	unsigned long long result;
+};
+
+static const struct cmdline_test_memparse_entry testdata[] = {
+	{ "0",				"",	0ULL },
+	{ "1",				"",	1ULL },
+	{ "a",				"a",	0ULL },
+	{ "k",				"k",	0ULL },
+	{ "E",				"E",	0ULL },
+	{ "0xb",			"",	11ULL },
+	{ "0xz",			"x",	0ULL },
+	{ "1234",			"",	1234ULL },
+	{ "04567",			"",	2423ULL },
+	{ "0x9876",			"",	39030LL },
+	{ "05678",			"8",	375ULL },
+	{ "0xabcdefz",			"z",	11259375ULL },
+	{ "0cdba",			"c",	0ULL },
+	{ "4K",				"",	SZ_4K },
+	{ "0x10k@0xaaaabbbb",		"@",	SZ_16K },
+	{ "32M",			"",	SZ_32M },
+	{ "067m:foo",			":",	55 * SZ_1M },
+	{ "2G;bar=baz",			";",	SZ_2G },
+	{ "07gz",			"z",	7ULL * SZ_1G },
+	{ "3T+data",			"+",	3 * SZ_1T },
+	{ "04t,ro",			",",	SZ_4T },
+	{ "012p",			"",	11258999068426240ULL },
+	{ "7P,sync",			",",	7881299347898368ULL },
+	{ "0x2e",			"",	46ULL },
+	{ "2E and more",		" ",	2305843009213693952ULL },
+	{ "18446744073709551615",	"",	ULLONG_MAX },
+	{ "0xffffffffffffffff0",	"",	ULLONG_MAX },
+	{ "1111111111111111111T",	"",	ULLONG_MAX },
+	{ "222222222222222222222G",	"",	ULLONG_MAX },
+	{ "3333333333333333333333M",	"",	ULLONG_MAX },
+};
+
+static void cmdline_test_memparse(struct kunit *test)
+{
+	const struct cmdline_test_memparse_entry *e;
+	unsigned long long ret;
+	char *retptr;
+
+	for (e = testdata; e < testdata + ARRAY_SIZE(testdata); e++) {
+		ret = memparse(e->input, &retptr);
+		KUNIT_EXPECT_EQ_MSG(test, ret, e->result,
+				    "    when parsing '%s'", e->input);
+		KUNIT_EXPECT_EQ_MSG(test, *retptr, *e->unrecognized,
+				    "    when parsing '%s'", e->input);
+	}
+}
+
 static struct kunit_case cmdline_test_cases[] = {
 	KUNIT_CASE(cmdline_test_noint),
 	KUNIT_CASE(cmdline_test_lead_int),
 	KUNIT_CASE(cmdline_test_tail_int),
 	KUNIT_CASE(cmdline_test_range),
+	KUNIT_CASE(cmdline_test_memparse),
 	{}
 };
 
-- 
2.53.0


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

[PATCH v9 5/6] lib/cmdline: adjust a few comments to fix kernel-doc -Wreturn warnings

[Dmitry Antipov]

Fix 'get_option()', 'memparse()' and 'parse_option_str()' comments
to match the commonly used style as suggested by kernel-doc -Wreturn.

Suggested-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
---
v2 and upwards: bump version to match the series
---
 lib/cmdline.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/lib/cmdline.c b/lib/cmdline.c
index f6e4b113ca9f..16cce6621cec 100644
--- a/lib/cmdline.c
+++ b/lib/cmdline.c
@@ -43,7 +43,7 @@ static int get_range(char **str, int *pint, int n)
  *	When @pint is NULL the function can be used as a validator of
  *	the current option in the string.
  *
- *	Return values:
+ *	Return:
  *	0 - no int in string
  *	1 - int found, no subsequent comma
  *	2 - int found including a subsequent comma
@@ -145,6 +145,9 @@ EXPORT_SYMBOL(get_options);
  *
  *	Parses a string into a number.  The number stored at @ptr is
  *	potentially suffixed with K, M, G, T, P, E.
+ *
+ *	Return: The value as recognized by simple_strtoull() multiplied
+ *	by the value as specified by suffix, if any.
  */
 
 unsigned long long memparse(const char *ptr, char **retptr)
@@ -205,7 +208,7 @@ EXPORT_SYMBOL(memparse);
  *	This function parses a string containing a comma-separated list of
  *	strings like a=b,c.
  *
- *	Return true if there's such option in the string, or return false.
+ *	Return: True if there's such option in the string or false otherwise.
  */
 bool parse_option_str(const char *str, const char *option)
 {
-- 
2.53.0


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

[PATCH v9 6/6] riscv: export symbols needed for riscv32 EFI stub

[Dmitry Antipov]

Fix the following link error observed when building EFI-enabled
kernel with CONFIG_EFI_STUB=y and CONFIG_EFI_GENERIC_STUB=y:

riscv32-linux-gnu-ld: ./drivers/firmware/efi/libstub/lib-cmdline.stub.o: in function `__efistub_.L49':
__efistub_cmdline.c:(.init.text+0x1f2): undefined reference to `__efistub___ashldi3'
riscv32-linux-gnu-ld: __efistub_cmdline.c:(.init.text+0x202): undefined reference to `__efistub___lshrdi3'

Note: compile (gcc 15.2.1 and clang 21.1.8) tested only.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202603041925.KLKqpK6N-lkp@intel.com
Suggested-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
---
v9: initial version to join the series
---
 arch/riscv/kernel/image-vars.h | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/arch/riscv/kernel/image-vars.h b/arch/riscv/kernel/image-vars.h
index 3bd9d06a8b8f..888bb1ddf945 100644
--- a/arch/riscv/kernel/image-vars.h
+++ b/arch/riscv/kernel/image-vars.h
@@ -32,6 +32,15 @@ __efistub___init_text_end	= __init_text_end;
 __efistub_sysfb_primary_display	= sysfb_primary_display;
 #endif
 
+#ifdef CONFIG_CC_IS_GCC
+/*
+ * Double-word integer shifts are used by the library code and so EFI stub as
+ * well. Not needed for clang and please let me know if anyone understands why.
+ */
+PROVIDE(__efistub___lshrdi3    = __lshrdi3);
+PROVIDE(__efistub___ashldi3    = __ashldi3);
+#endif /* CONFIG_CC_IS_GCC */
+
 #endif
 
 #endif /* __RISCV_KERNEL_IMAGE_VARS_H */
-- 
2.53.0


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

Re: [PATCH v9 0/6] lib and lib/cmdline enhancements

[Andrew Morton]

On Fri,  3 Apr 2026 13:33:32 +0300 Dmitry Antipov <dmantipov@yandex.ru> wrote:

> Adjust '_parse_integer_limit()' and 'memparse()' to not ignore overflows,
> extend string to 64-bit integer conversion tests, add KUnit-based test
> for 'memparse()', fix kernel-doc glitches found in lib/cmdline.c and
> adjust riscv32 linker script to export symbols needed for EFI stub.

Thanks, Dmitry.

I've updated mm.git's mm-nonmm-unstable with this version.  I'll aim to
upstream this series during the upcoming merge window.

Here's how the v9 series altered mm.git:

 arch/riscv/kernel/image-vars.h |    9 +++++++++
 1 file changed, 9 insertions(+)

--- a/arch/riscv/kernel/image-vars.h~b
+++ a/arch/riscv/kernel/image-vars.h
@@ -32,6 +32,15 @@ __efistub___init_text_end	= __init_text_
 __efistub_sysfb_primary_display	= sysfb_primary_display;
 #endif
 
+#ifdef CONFIG_CC_IS_GCC
+/*
+ * Double-word integer shifts are used by the library code and so EFI stub as
+ * well. Not needed for clang and please let me know if anyone understands why.
+ */
+PROVIDE(__efistub___lshrdi3    = __lshrdi3);
+PROVIDE(__efistub___ashldi3    = __ashldi3);
+#endif /* CONFIG_CC_IS_GCC */
+
 #endif
 
 #endif /* __RISCV_KERNEL_IMAGE_VARS_H */
_


_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

Re: [PATCH v9 0/6] lib and lib/cmdline enhancements

[Andrew Morton]

On Fri,  3 Apr 2026 13:33:32 +0300 Dmitry Antipov <dmantipov@yandex.ru> wrote:

> Adjust '_parse_integer_limit()' and 'memparse()' to not ignore overflows,
> extend string to 64-bit integer conversion tests, add KUnit-based test
> for 'memparse()', fix kernel-doc glitches found in lib/cmdline.c and
> adjust riscv32 linker script to export symbols needed for EFI stub.

AI review asked a few questions:
https://sashiko.dev/#/patchset/20260403103338.1122415-1-dmantipov@yandex.ru



_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

Re: [PATCH v9 0/6] lib and lib/cmdline enhancements

[Dmitry Antipov]

On Fri, 2026-04-03 at 12:25 -0700, Andrew Morton wrote:

> AI review asked a few questions:
> https://sashiko.dev/#/patchset/20260403103338.1122415-1-dmantipov@yandex.ru

Hmm....is AI the only "person" interesting in what's going on here?

Is https://lists.infradead.org/pipermail/linux-riscv/2021-June/007376.html still true?

Are there any chances to see how it works under QEMU? AFAICS riscv64
defconfig boots just out-of-the-box with qemu-system-riscv64 -M virt,
but riscv32 shows nothing beyond SBI messages.

Dmitry

_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

Re: [PATCH v9 0/6] lib and lib/cmdline enhancements

[Andy Shevchenko]

On Sun, Apr 05, 2026 at 12:13:04PM +0300, Dmitry Antipov wrote:
> On Fri, 2026-04-03 at 12:25 -0700, Andrew Morton wrote:
> 
> > AI review asked a few questions:
> > https://sashiko.dev/#/patchset/20260403103338.1122415-1-dmantipov@yandex.ru
> 
> Hmm....is AI the only "person" interesting in what's going on here?
> 
> Is https://lists.infradead.org/pipermail/linux-riscv/2021-June/007376.html still true?

Not sure, we probably need to Cc more Risc-V people. But don't ask me who they
should be, perhaps MAINTAINERS can give a good guess(es).

> Are there any chances to see how it works under QEMU? AFAICS riscv64
> defconfig boots just out-of-the-box with qemu-system-riscv64 -M virt,
> but riscv32 shows nothing beyond SBI messages.

Reading all this I think we may leave proposed solution (by Ard) and
if one wants to fix it in the future, it would be still doable.

Another option is to postpone memparse() patch and submit it after
v7.1-rc1 is out.

-- 
With Best Regards,
Andy Shevchenko



_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

Re: [PATCH v9 6/6] riscv: export symbols needed for riscv32 EFI stub

[Charlie Jenkins]

On Fri, Apr 03, 2026 at 01:33:38PM +0300, Dmitry Antipov wrote:
> Fix the following link error observed when building EFI-enabled
> kernel with CONFIG_EFI_STUB=y and CONFIG_EFI_GENERIC_STUB=y:
> 
> riscv32-linux-gnu-ld: ./drivers/firmware/efi/libstub/lib-cmdline.stub.o: in function `__efistub_.L49':
> __efistub_cmdline.c:(.init.text+0x1f2): undefined reference to `__efistub___ashldi3'
> riscv32-linux-gnu-ld: __efistub_cmdline.c:(.init.text+0x202): undefined reference to `__efistub___lshrdi3'
> 
> Note: compile (gcc 15.2.1 and clang 21.1.8) tested only.

I attempted to boot this on riscv32 in qemu with u-boot. I used
CONFIG_EFI_ZBOOT to generate vmlinux.efi. With this option enabled,
there is the same undefined reference errors with this patch applied.
I have not found a good way to resolve this issue, so I resorted to
copying the __lshrdi3() implementation over into
drivers/firmware/efi/libstub/riscv.c and that successfully booted with
EFI.

- Charlie

> 
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202603041925.KLKqpK6N-lkp@intel.com
> Suggested-by: Ard Biesheuvel <ardb@kernel.org>
> Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
> ---
> v9: initial version to join the series
> ---
>  arch/riscv/kernel/image-vars.h | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/arch/riscv/kernel/image-vars.h b/arch/riscv/kernel/image-vars.h
> index 3bd9d06a8b8f..888bb1ddf945 100644
> --- a/arch/riscv/kernel/image-vars.h
> +++ b/arch/riscv/kernel/image-vars.h
> @@ -32,6 +32,15 @@ __efistub___init_text_end	= __init_text_end;
>  __efistub_sysfb_primary_display	= sysfb_primary_display;
>  #endif
>  
> +#ifdef CONFIG_CC_IS_GCC
> +/*
> + * Double-word integer shifts are used by the library code and so EFI stub as
> + * well. Not needed for clang and please let me know if anyone understands why.
> + */
> +PROVIDE(__efistub___lshrdi3    = __lshrdi3);
> +PROVIDE(__efistub___ashldi3    = __ashldi3);
> +#endif /* CONFIG_CC_IS_GCC */
> +
>  #endif
>  
>  #endif /* __RISCV_KERNEL_IMAGE_VARS_H */
> -- 
> 2.53.0
> 
> 
> _______________________________________________
> linux-riscv mailing list
> linux-riscv@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv
> 

_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

Re: [PATCH v9 6/6] riscv: export symbols needed for riscv32 EFI stub

[Andy Shevchenko]

On Wed, Apr 08, 2026 at 10:00:18PM -0700, Charlie Jenkins wrote:
> On Fri, Apr 03, 2026 at 01:33:38PM +0300, Dmitry Antipov wrote:
> > Fix the following link error observed when building EFI-enabled
> > kernel with CONFIG_EFI_STUB=y and CONFIG_EFI_GENERIC_STUB=y:
> > 
> > riscv32-linux-gnu-ld: ./drivers/firmware/efi/libstub/lib-cmdline.stub.o: in function `__efistub_.L49':
> > __efistub_cmdline.c:(.init.text+0x1f2): undefined reference to `__efistub___ashldi3'
> > riscv32-linux-gnu-ld: __efistub_cmdline.c:(.init.text+0x202): undefined reference to `__efistub___lshrdi3'
> > 
> > Note: compile (gcc 15.2.1 and clang 21.1.8) tested only.
> 
> I attempted to boot this on riscv32 in qemu with u-boot. I used
> CONFIG_EFI_ZBOOT to generate vmlinux.efi. With this option enabled,
> there is the same undefined reference errors with this patch applied.
> I have not found a good way to resolve this issue, so I resorted to
> copying the __lshrdi3() implementation over into
> drivers/firmware/efi/libstub/riscv.c and that successfully booted with
> EFI.

Thank you for trying and testing this.
Can you instead this patch take the mini-series from 20260409033824.465455-1-dmantipov@yandex.ru?

> > Closes: https://lore.kernel.org/oe-kbuild-all/202603041925.KLKqpK6N-lkp@intel.com

-- 
With Best Regards,
Andy Shevchenko



_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv