Re: [SMP lock BUG?] Re: Feedback on preemptible kernel patch

[Manfred Spraul]

This is a multi-part message in MIME format.

------=_NextPart_000_0013_01C13913.2720D6F0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

> #define kmap_atomic(page,idx) ctx_sw_off(); kmap(page);
> #define kunmap_atomic(page,idx) ctx_sw_on(); kunmap(page);
>
No. kmap_atomic is called from interrupt context, and kmap calls
schedule().

I thought about the attached patch (completely untested).

--
    Manfred

------=_NextPart_000_0013_01C13913.2720D6F0
Content-Type: application/octet-stream;
	name="patch-untested"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
	filename="patch-untested"

--- highmem.h.prev	Sun Sep  9 08:59:04 2001=0A=
+++ highmem.h	Sun Sep  9 09:00:07 2001=0A=
@@ -88,6 +88,7 @@=0A=
 	if (page < highmem_start_page)=0A=
 		return page_address(page);=0A=
 =0A=
+	ctx_sw_off();=0A=
 	idx =3D type + KM_TYPE_NR*smp_processor_id();=0A=
 	vaddr =3D __fix_to_virt(FIX_KMAP_BEGIN + idx);=0A=
 #if HIGHMEM_DEBUG=0A=
@@ -119,6 +120,7 @@=0A=
 	pte_clear(kmap_pte-idx);=0A=
 	__flush_tlb_one(vaddr);=0A=
 #endif=0A=
+	ctx_sw_on();=0A=
 }=0A=
 =0A=
 #endif /* __KERNEL__ */=0A=

------=_NextPart_000_0013_01C13913.2720D6F0--