AuroraMiddleware/harfbuzz
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

[ENOMEM] handle allocation failures during page compaction.

Browse Source
This commit is contained in:
Garret Rieger 2020-09-16 13:56:33 -07:00
parent a4c3732f59
commit 0ba73e980c
1 changed files with 39 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

50
src/hb-set.hh
View File

@ -389,6 +389,11 @@ struct hb_set_t
{
if (ds <= de)
{
// Pre-allocate the workspace that compact() will need so we can bail on allocation failure
// before attempting to rewrite the page map.
hb_vector_t<uint32_t> compact_workspace;
if (unlikely (!allocate_compact_workspace (&compact_workspace))) return;
unsigned int write_index = 0;
for (unsigned int i = 0; i < page_map.length; i++)
{
@ -396,11 +401,12 @@ struct hb_set_t
if (m < ds || de < m)
page_map[write_index++] = page_map[i];
}
compact (write_index);
compact (&compact_workspace, write_index);
resize (write_index);
}
}
public:
void del_range (hb_codepoint_t a, hb_codepoint_t b)
{
@ -512,18 +518,35 @@ struct hb_set_t
return true;
}
void compact (unsigned int length)
bool allocate_compact_workspace(hb_vector_t<uint32_t>* workspace)
{
hb_vector_t<uint32_t> old_index_to_page_map_index;
// TODO(grieger): need to handle resize failure here.
old_index_to_page_map_index.resize(pages.length);
for (uint32_t i = 0; i < old_index_to_page_map_index.length; i++)
old_index_to_page_map_index[i] = 0xFFFFFFFF;
if (unlikely(!workspace->resize (pages.length)))
{
successful = false;
return false;
}
return true;
}
/*
* workspace should be a pre-sized vector allocated to hold at exactly pages.length
* elements.
*/
void compact (hb_vector_t<uint32_t>* workspace,
unsigned int length)
{
assert(workspace->length == pages.length);
hb_vector_t<uint32_t>* old_index_to_page_map_index = workspace;
for (uint32_t i = 0; i < old_index_to_page_map_index->length; i++)
(*old_index_to_page_map_index)[i] = 0xFFFFFFFF;
for (uint32_t i = 0; i < length; i++)
old_index_to_page_map_index[page_map[i].index] = i;
(*old_index_to_page_map_index)[page_map[i].index] = i;
compact_pages (old_index_to_page_map_index);
compact_pages (*old_index_to_page_map_index);
}
void compact_pages (const hb_vector_t<uint32_t>& old_index_to_page_map_index)
@ -555,6 +578,12 @@ struct hb_set_t
unsigned int count = 0, newCount = 0;
unsigned int a = 0, b = 0;
unsigned int write_index = 0;
// Pre-allocate the workspace that compact() will need so we can bail on allocation failure
// before attempting to rewrite the page map.
hb_vector_t<uint32_t> compact_workspace;
if (!Op::passthru_left && unlikely (!allocate_compact_workspace (&compact_workspace))) return;
for (; a < na && b < nb; )
{
if (page_map[a].major == other->page_map[b].major)
@ -596,11 +625,10 @@ struct hb_set_t
{
na = write_index;
next_page = write_index;
compact (write_index);
compact (&compact_workspace, write_index);
}
if (!resize (count))
// TODO(grieger): does this leave us in an invalid state?
return;
newCount = count;