iio: buffer: allocation and freeing buffers fix and optimization
v2 cover iio: buffer: allocation and freeing buffers fix and optimization
Yang submitted a fix, but I think the code can be refactored a bit to be more robust against similar mistakes in the future, if any.
In v2:
- put SoB Yang's patch (it's good for backporting)
- added refactoring patch on top of Yang's fix
Andy Shevchenko (1):
iio: buffer: Use dedicated variable in
iio_buffers_alloc_sysfs_and_mask()
Yang Yingliang (1):
iio: buffer: Fix double-free in iio_buffers_alloc_sysfs_and_mask()
drivers/iio/industrialio-buffer.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-)
Signature · 1 line
2.33.0
PATCH v2 1/2 iio: buffer: Fix double-free in iio_buffers_alloc_sysfs_and_mask()
From: Yang Yingliang <[email protected]>
When __iio_buffer_alloc_sysfs_and_mask() failed, 'unwind_idx' should be set to 'i - 1' to prevent double-free when cleanup resources.
BUG: KASAN: double-free or invalid-free in __iio_buffer_free_sysfs_and_mask+0x32/0xb0 [industrialio] Call Trace: kfree+0x117/0x4c0 __iio_buffer_free_sysfs_and_mask+0x32/0xb0 [industrialio] iio_buffers_alloc_sysfs_and_mask+0x60d/0x1570 [industrialio] __iio_device_register+0x483/0x1a30 [industrialio] ina2xx_probe+0x625/0x980 [ina2xx_adc]
--- drivers/iio/industrialio-buffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/industrialio-buffer.c +1 -1
@@ -1616,7 +1616,7 @@ int iio_buffers_alloc_sysfs_and_mask(struct iio_dev *indio_dev) buffer = iio_dev_opaque->attached_buffers[i]; ret =__iio_buffer_alloc_sysfs_and_mask (buffer, indio_dev, i);if (ret) { - unwind_idx = i; + unwind_idx = i -1 ; goto error_unwind_sysfs_and_mask; } }
Signature · 1 line
2.33.0
PATCH v2 2/2 iio: buffer: Use dedicated variable in iio_buffers_alloc_sysfs_and_mask()
In reply to Andy ShevchenkoUse dedicated variable for index in the loop in the iio_buffers_alloc_sysfs_and_mask(). This will make code cleaner and less error prone as proved by previous changes done in this function.
--- drivers/iio/industrialio-buffer.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-)
drivers/iio/industrialio-buffer.c +7 -11
@@ -1596,8 +1596,7 @@ int iio_buffers_alloc_sysfs_and_mask(struct iio_dev *indio_dev)struct iio_dev_opaque *iio_dev_opaque =to_iio_dev_opaque (indio_dev);const struct iio_chan_spec *channels;struct iio_buffer *buffer; -int unwind_idx; -int ret, i; +int ret, i, idx;size_t sz; channels = indio_dev->channels;
@@ -1612,15 +1611,12 @@ int iio_buffers_alloc_sysfs_and_mask(struct iio_dev *indio_dev)if (!iio_dev_opaque->attached_buffers_cnt)return 0 ; -for (i =0 ; i < iio_dev_opaque->attached_buffers_cnt; i++) { - buffer = iio_dev_opaque->attached_buffers[i]; - ret =__iio_buffer_alloc_sysfs_and_mask (buffer, indio_dev, i); -if (ret) { - unwind_idx = i -1 ; +for (idx =0 ; idx < iio_dev_opaque->attached_buffers_cnt; idx++) { + buffer = iio_dev_opaque->attached_buffers[idx]; + ret =__iio_buffer_alloc_sysfs_and_mask (buffer, indio_dev, idx); +if (ret) goto error_unwind_sysfs_and_mask; - } } - unwind_idx = iio_dev_opaque->attached_buffers_cnt -1 ; sz =sizeof (*(iio_dev_opaque->buffer_ioctl_handler)); iio_dev_opaque->buffer_ioctl_handler =kzalloc (sz,GFP_KERNEL );
@@ -1636,8 +1632,8 @@ int iio_buffers_alloc_sysfs_and_mask(struct iio_dev *indio_dev)return 0 ; error_unwind_sysfs_and_mask: -for (; unwind_idx >=0 ; unwind_idx--) { - buffer = iio_dev_opaque->attached_buffers[unwind_idx]; +while (idx--) { + buffer = iio_dev_opaque->attached_buffers[idx];__iio_buffer_free_sysfs_and_mask (buffer); }return ret;
Signature · 1 line
2.33.0
Jonathan Cameron
maintainerQuoted 11 lines · On Wed, 13 Oct 2021 12:49:21 +0300 Andy Shevchenko…
Yang submitted a fix, but I think the code can be refactored a bit to be more robust against similar mistakes in the future, if any.
In v2:
- put SoB Yang's patch (it's good for backporting)
- added refactoring patch on top of Yang's fixAndy Shevchenko (1):
iio: buffer: Use dedicated variable in
iio_buffers_alloc_sysfs_and_mask()
Quoted 6 lines
Yang Yingliang (1):
iio: buffer: Fix double-free iniio_buffers_alloc_sysfs_and_mask()drivers/iio/industrialio-buffer.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-)
1st patch applied to the fixes-togreg branch of iio.git. I may well end up sending these in the merge window anyway in which case I'll probably stick patch 2 on top of it before sending. If not I'll pick that up next cycle now.
Thanks,
Jonathan
Quoted 17 lines · Jonathan Cameron
Quoted 14 lines · On Wed, 13 Oct 2021 12:49:21 +0300 Andy Shevchenko…
Yang submitted a fix, but I think the code can be refactored a bit to be more robust against similar mistakes in the future, if any.
In v2:
- put SoB Yang's patch (it's good for backporting)
- added refactoring patch on top of Yang's fixAndy Shevchenko (1):
iio: buffer: Use dedicated variable in
iio_buffers_alloc_sysfs_and_mask()Yang Yingliang (1):
iio: buffer: Fix double-free iniio_buffers_alloc_sysfs_and_mask()
Quoted 6 lines
Quoted 3 lines
drivers/iio/industrialio-buffer.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-)1st patch applied to the fixes-togreg branch of iio.git. I may well end up sending these in the merge window anyway in which case I'll probably stick patch 2 on top of it before sending. If not I'll pick that up next cycle now.
Is it a right time now?
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 18 lines · On Mon, 15 Nov 2021 13:12:24 +0200 Andy Shevchenko…
Quoted 17 lines · Jonathan Cameron
Quoted 14 lines · On Wed, 13 Oct 2021 12:49:21 +0300 Andy Shevchenko…
Yang submitted a fix, but I think the code can be refactored a bit to be more robust against similar mistakes in the future, if any.
In v2:
- put SoB Yang's patch (it's good for backporting)
- added refactoring patch on top of Yang's fixAndy Shevchenko (1):
iio: buffer: Use dedicated variable in
iio_buffers_alloc_sysfs_and_mask()Yang Yingliang (1):
iio: buffer: Fix double-free iniio_buffers_alloc_sysfs_and_mask()
Quoted 9 lines
Quoted 6 lines
Quoted 3 lines
drivers/iio/industrialio-buffer.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-)1st patch applied to the fixes-togreg branch of iio.git. I may well end up sending these in the merge window anyway in which case I'll probably stick patch 2 on top of it before sending. If not I'll pick that up next cycle now.
Is it a right time now?
Applied, but needed a bit of hand tweaking as patches have crossed with this.
Pushed out as testing for 0-day to see if we missed anything.
Thanks,
Jonathan