On 9/10/26 03:24, Jamin Lin wrote:
> The ACRY RSA qtest allocates two ACRY_SRAM_SIZE buffers on the
> stack. Since each buffer is nearly 9 KiB, this unnecessarily increases
> the stack usage of test_modexp_rsa() and may cause unexpected failures
> on environments with a limited stack size.
>
> This reduces the stack usage while preserving the existing test
> behavior.
>
> Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
> ---
> tests/qtest/aspeed-acry-test.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/tests/qtest/aspeed-acry-test.c b/tests/qtest/aspeed-acry-test.c
> index ec348c2206..51048d697c 100644
> --- a/tests/qtest/aspeed-acry-test.c
> +++ b/tests/qtest/aspeed-acry-test.c
> @@ -327,8 +327,8 @@ static void test_modexp_rsa(const void *opaque)
> .alg = QCRYPTO_AK_CIPHER_ALGO_RSA,
> .u.rsa.padding_alg = QCRYPTO_RSA_PADDING_ALGO_RAW,
> };
> - uint8_t dram_buf[ACRY_SRAM_SIZE] = { 0 };
> - uint8_t sram_buf[ACRY_SRAM_SIZE] = { 0 };
> + g_autofree uint8_t *dram_buf = g_malloc0(ACRY_SRAM_SIZE);
> + g_autofree uint8_t *sram_buf = g_malloc0(ACRY_SRAM_SIZE);
> uint8_t result[ACRY_MAX_BYTES] = { 0 };
> QTestState *qts;
>
> @@ -345,7 +345,7 @@ static void test_modexp_rsa(const void *opaque)
> put_bignum_be_bytes(dram_buf, ACRY_MOD_OFFSET, t->n, t->n_len);
> put_bignum_be_bytes(dram_buf, ACRY_DATA_OFFSET, t->m, t->m_len);
>
> - qtest_memwrite(qts, c->dram_addr, dram_buf, sizeof(dram_buf));
> + qtest_memwrite(qts, c->dram_addr, dram_buf, ACRY_SRAM_SIZE);
>
> qtest_writel(qts, c->acry_addr + ACRY_DMA_CMD, ACRY_DMA_CMD_DMEM_AHB);
> qtest_writel(qts, c->acry_addr + ACRY_DMA_SRC, c->dram_addr);
> @@ -362,7 +362,7 @@ static void test_modexp_rsa(const void *opaque)
> g_assert_cmphex(qtest_readl(qts, c->acry_addr + ACRY_STATUS), ==,
> ACRY_STATUS_RSA_DONE);
>
> - qtest_memread(qts, c->sram_addr, sram_buf, sizeof(sram_buf));
> + qtest_memread(qts, c->sram_addr, sram_buf, ACRY_SRAM_SIZE);
> get_bignum_be_bytes(sram_buf, ACRY_DATA_OFFSET, result, t->c_len);
> g_assert_cmpmem(result, t->c_len, t->c, t->c_len);
>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Thanks,
C.