Skip to contentSkip to conversation

x86/sgx: Initialize "ret" in sgx_ioc_enclave_add_pages()

PATCH In review Jarkko Sakkinen 3 participants 3 messages Dec 2, 2020 – Dec 3, 2020

Subsystems (2)
lore
Share & tools
lore

Jarkko Sakkinen

author maintainer
#0 lore raw

PATCH 1/1 x86/sgx: Initialize "ret" in sgx_ioc_enclave_add_pages()

Initialize "ret" to zero as otherwise a zero length address range will leave it uninitialized.

Cc Dave Hansen <[email protected]>
Cc Borislav Petkov <[email protected]>
Fixes c6d26d370767 ("x86/sgx: Add SGX_IOC_ENCLAVE_ADD_PAGES")
---
 arch/x86/kernel/cpu/sgx/ioctl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/sgx/ioctl.c +1 -1
index c206aee80a04..36a244504da2 100644
@@ -415,7 +415,7 @@ static long sgx_ioc_enclave_add_pages(struct sgx_encl *encl, void __user *arg)
 	struct sgx_enclave_add_pages add_arg;
 	struct sgx_secinfo secinfo;
 	unsigned long c;
-	int ret;
+	int ret = 0;
 
 	if (!test_bit(SGX_ENCL_CREATED, &encl->flags) ||
 	    test_bit(SGX_ENCL_INITIALIZED, &encl->flags))
Signature · 1 line
2.27.0
maintainer
#1 lore raw
Quoted 2 lines · Jarkko Sakkinen

Initialize "ret" to zero as otherwise a zero length address range will leave it uninitialized.

That length is:

* @length: length of the data (multiple of the page size)

I think we wanna fail this even earlier when it wants to add data of length 0 because that sounds nonsensical to me. Or is there some use case for zero-length data?

Btw, pls CC lkml on patch submissions.

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
#2 lore raw
Quoted 6 lines · Borislav Petkov
Quoted 2 lines · Jarkko Sakkinen

Initialize "ret" to zero as otherwise a zero length address range will leave it uninitialized.

That length is:

Quoted 7 lines

* @length: length of the data (multiple of the page size)

I think we wanna fail this even earlier when it wants to add data of length 0 because that sounds nonsensical to me. Or is there some use case for zero-length data?

Btw, pls CC lkml on patch submissions.

Yeah, makes sense to fail right after copying the parameter struct with -EINVAL.

Quoted 8 lines

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette

/Jarkko