From: popcornmix <popcornmix@gmail.com>
Date: Tue, 2 May 2017 15:36:05 +0000 (+0100)
Subject: vcsm: Treat EBUSY as success rather than SIGBUS
X-Git-Tag: archive/raspbian/4.9.82-1+deb9u3+rpi1_jessie~5^2~241
X-Git-Url: https://dgit.raspbian.org/%22http://www.example.com/cgi/%22/%22http:/www.example.com/cgi/%22?a=commitdiff_plain;h=ff7721467a7c8a1fd1b1790e66e8bafd907bdc30;p=linux-4.9.git

vcsm: Treat EBUSY as success rather than SIGBUS

Currently if two cores access the same page concurrently one will return VM_FAULT_NOPAGE
and the other VM_FAULT_SIGBUS crashing the user code.

Also report when mapping fails.

Signed-off-by: popcornmix <popcornmix@gmail.com>
---

diff --git a/drivers/char/broadcom/vc_sm/vmcs_sm.c b/drivers/char/broadcom/vc_sm/vmcs_sm.c
index 1db6716c2c0c..ee4e05948c43 100644
--- a/drivers/char/broadcom/vc_sm/vmcs_sm.c
+++ b/drivers/char/broadcom/vc_sm/vmcs_sm.c
@@ -1181,11 +1181,20 @@ static int vcsm_vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	switch (ret) {
 	case 0:
 	case -ERESTARTSYS:
+	/*
+	* EBUSY is ok: this just means that another thread
+	* already did the job.
+	*/
+	case -EBUSY:
 		return VM_FAULT_NOPAGE;
 	case -ENOMEM:
 	case -EAGAIN:
+		pr_err("[%s]: failed to map page pfn:%lx virt:%lx ret:%d\n", __func__,
+			pfn, (unsigned long)vmf->virtual_address, ret);
 		return VM_FAULT_OOM;
 	default:
+		pr_err("[%s]: failed to map page pfn:%lx virt:%lx ret:%d\n", __func__,
+			pfn, (unsigned long)vmf->virtual_address, ret);
 		return VM_FAULT_SIGBUS;
 	}
 }