| Message ID | 20210514172247.176750-7-david@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | fs/proc/kcore: don't read offline sections, logically offline pages and hwpoisoned pages | expand |
On Fri, May 14, 2021 at 07:22:47PM +0200, David Hildenbrand wrote: > Let's properly synchronize with drivers that set PageOffline(). > Unfreeze/thaw every now and then, so drivers that want to set PageOffline() > can make progress. > > Signed-off-by: David Hildenbrand <david@redhat.com> Acked-by: Mike Rapoport <rppt@linux.ibm.com> > --- > fs/proc/kcore.c | 13 +++++++++++++ > 1 file changed, 13 insertions(+) > > diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c > index 92ff1e4436cb..982e694aae77 100644 > --- a/fs/proc/kcore.c > +++ b/fs/proc/kcore.c > @@ -313,6 +313,7 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) > { > char *buf = file->private_data; > size_t phdrs_offset, notes_offset, data_offset; > + size_t page_offline_frozen = 1; > size_t phdrs_len, notes_len; > struct kcore_list *m; > size_t tsz; > @@ -322,6 +323,11 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) > int ret = 0; > > down_read(&kclist_lock); > + /* > + * Don't race against drivers that set PageOffline() and expect no > + * further page access. > + */ > + page_offline_freeze(); > > get_kcore_size(&nphdr, &phdrs_len, ¬es_len, &data_offset); > phdrs_offset = sizeof(struct elfhdr); > @@ -480,6 +486,12 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) > } > } > > + if (page_offline_frozen++ % MAX_ORDER_NR_PAGES == 0) { > + page_offline_thaw(); > + cond_resched(); > + page_offline_freeze(); > + } > + > if (&m->list == &kclist_head) { > if (clear_user(buffer, tsz)) { > ret = -EFAULT; > @@ -565,6 +577,7 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) > } > > out: > + page_offline_thaw(); > up_read(&kclist_lock); > if (ret) > return ret; > -- > 2.31.1 >
On Fri, May 14, 2021 at 07:22:47PM +0200, David Hildenbrand wrote: > Let's properly synchronize with drivers that set PageOffline(). > Unfreeze/thaw every now and then, so drivers that want to set PageOffline() > can make progress. > > Signed-off-by: David Hildenbrand <david@redhat.com> Reviewed-by: Oscar Salvador <osalvador@suse.de>
diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index 92ff1e4436cb..982e694aae77 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -313,6 +313,7 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) { char *buf = file->private_data; size_t phdrs_offset, notes_offset, data_offset; + size_t page_offline_frozen = 1; size_t phdrs_len, notes_len; struct kcore_list *m; size_t tsz; @@ -322,6 +323,11 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) int ret = 0; down_read(&kclist_lock); + /* + * Don't race against drivers that set PageOffline() and expect no + * further page access. + */ + page_offline_freeze(); get_kcore_size(&nphdr, &phdrs_len, ¬es_len, &data_offset); phdrs_offset = sizeof(struct elfhdr); @@ -480,6 +486,12 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) } } + if (page_offline_frozen++ % MAX_ORDER_NR_PAGES == 0) { + page_offline_thaw(); + cond_resched(); + page_offline_freeze(); + } + if (&m->list == &kclist_head) { if (clear_user(buffer, tsz)) { ret = -EFAULT; @@ -565,6 +577,7 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) } out: + page_offline_thaw(); up_read(&kclist_lock); if (ret) return ret;
Let's properly synchronize with drivers that set PageOffline(). Unfreeze/thaw every now and then, so drivers that want to set PageOffline() can make progress. Signed-off-by: David Hildenbrand <david@redhat.com> --- fs/proc/kcore.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)