The following commit has been merged into the x86/urgent branch of tip:
Commit-ID: 6b7b71a70af6d75e0a9eddf4b01e4383a78b8a5e Gitweb: https://git.kernel.org/tip/6b7b71a70af6d75e0a9eddf4b01e4383a78b8a5e Author: Haitao Huang haitao.huang@linux.intel.com AuthorDate: Thu, 19 Oct 2023 19:53:53 -07:00 Committer: Ingo Molnar mingo@kernel.org CommitterDate: Fri, 20 Oct 2023 10:05:48 +02:00
x86/sgx: Return VM_FAULT_SIGBUS instead of VM_FAULT_OOM for EPC exhaustion
In the EAUG on page fault path, VM_FAULT_OOM is returned when the Enclave Page Cache (EPC) runs out. This may trigger unneeded OOM kill that will not free any EPCs. Return VM_FAULT_SIGBUS instead.
Fixes: 5a90d2c3f5ef ("x86/sgx: Support adding of pages to an initialized enclave") Signed-off-by: Haitao Huang haitao.huang@linux.intel.com Signed-off-by: Ingo Molnar mingo@kernel.org Cc: stable@vger.kernel.org # v6.0+ --- arch/x86/kernel/cpu/sgx/encl.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/cpu/sgx/encl.c b/arch/x86/kernel/cpu/sgx/encl.c index 279148e..d13b7e4 100644 --- a/arch/x86/kernel/cpu/sgx/encl.c +++ b/arch/x86/kernel/cpu/sgx/encl.c @@ -322,7 +322,7 @@ struct sgx_encl_page *sgx_encl_load_page(struct sgx_encl *encl, * ENCLS[EAUG] instruction. * * Returns: Appropriate vm_fault_t: VM_FAULT_NOPAGE when PTE was installed - * successfully, VM_FAULT_SIGBUS or VM_FAULT_OOM as error otherwise. + * successfully, VM_FAULT_SIGBUS as error otherwise. */ static vm_fault_t sgx_encl_eaug_page(struct vm_area_struct *vma, struct sgx_encl *encl, unsigned long addr) @@ -348,7 +348,7 @@ static vm_fault_t sgx_encl_eaug_page(struct vm_area_struct *vma, secinfo_flags = SGX_SECINFO_R | SGX_SECINFO_W | SGX_SECINFO_X; encl_page = sgx_encl_page_alloc(encl, addr - encl->base, secinfo_flags); if (IS_ERR(encl_page)) - return VM_FAULT_OOM; + return VM_FAULT_SIGBUS;
mutex_lock(&encl->lock);
This reverts commit 6b7b71a70af6d75e0a9eddf4b01e4383a78b8a5e. That patch was a mistake.
link: https://lore.kernel.org/all/op.2dfkbh2iwjvjmi@hhuan26-mobl.amr.corp.intel.co...
Signed-off-by: Haitao Huang haitao.huang@linux.intel.com --- Sorry for the trouble. Please use this to revert or drop it from the the tip/x86/urgent branch. --- arch/x86/kernel/cpu/sgx/encl.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/cpu/sgx/encl.c b/arch/x86/kernel/cpu/sgx/encl.c index d13b7e4ad0f5..279148e72459 100644 --- a/arch/x86/kernel/cpu/sgx/encl.c +++ b/arch/x86/kernel/cpu/sgx/encl.c @@ -322,7 +322,7 @@ struct sgx_encl_page *sgx_encl_load_page(struct sgx_encl *encl, * ENCLS[EAUG] instruction. * * Returns: Appropriate vm_fault_t: VM_FAULT_NOPAGE when PTE was installed - * successfully, VM_FAULT_SIGBUS as error otherwise. + * successfully, VM_FAULT_SIGBUS or VM_FAULT_OOM as error otherwise. */ static vm_fault_t sgx_encl_eaug_page(struct vm_area_struct *vma, struct sgx_encl *encl, unsigned long addr) @@ -348,7 +348,7 @@ static vm_fault_t sgx_encl_eaug_page(struct vm_area_struct *vma, secinfo_flags = SGX_SECINFO_R | SGX_SECINFO_W | SGX_SECINFO_X; encl_page = sgx_encl_page_alloc(encl, addr - encl->base, secinfo_flags); if (IS_ERR(encl_page)) - return VM_FAULT_SIGBUS; + return VM_FAULT_OOM;
mutex_lock(&encl->lock);
Hi,
Thanks for your patch.
FYI: kernel test robot notices the stable kernel rule is not satisfied.
The check is based on https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html#opti...
Rule: add the tag "Cc: stable@vger.kernel.org" in the sign-off area to have the patch automatically included in the stable tree. Subject: [PATCH] Revert "x86/sgx: Return VM_FAULT_SIGBUS instead of VM_FAULT_OOM for EPC exhaustion" Link: https://lore.kernel.org/stable/20231026225528.5738-1-haitao.huang%40linux.in...
On 10/26/23 15:55, Haitao Huang wrote:
This reverts commit 6b7b71a70af6d75e0a9eddf4b01e4383a78b8a5e. That patch was a mistake. link: https://lore.kernel.org/all/op.2dfkbh2iwjvjmi@hhuan26-mobl.amr.corp.intel.co...
Thanks for the patch. There was only one trivial commit on top, so I just dropped your commit and effectively rebased the other single one instead of applying a revert.
linux-stable-mirror@lists.linaro.org