Message ID | 20221118185118.1190044-1-michael@walle.cc |
---|---|
Headers | show |
Series | nvmem: core: introduce NVMEM layouts | expand |
On Fri, Nov 18, 2022 at 07:51:17PM +0100, Michael Walle wrote: > From: Miquel Raynal <miquel.raynal@bootlin.com> > > This layout applies no top of any non volatile storage device containing s/no/on/ > + table_len = hdr_len + data_len; > + if (table_len > ONIE_TLV_MAX_LEN) { > + dev_err(dev, "Invalid ONIE TLV data length\n"); > + return -EINVAL; > + } > + > + table = devm_kmalloc(dev, table_len, GFP_KERNEL); > + if (!table) > + return -ENOMEM; > + > + ret = nvmem_device_read(nvmem, 0, table_len, table); > + if (ret != table_len) > + goto free_data_buf; > + > + if (!onie_tlv_crc_is_valid(dev, table_len, table)) { > + ret = -EINVAL; > + goto free_data_buf; > + } > + > + data = table + hdr_len; > + ret = onie_tlv_add_cells(dev, nvmem, data_len, data); > + if (ret) > + goto free_data_buf; > + > +free_data_buf: > + kfree(table); This is allocated with devm_kmalloc(), you shouldn't free it here. Sascha
On Fri, Nov 18, 2022 at 07:51:06PM +0100, Michael Walle wrote: > If nvmem_add_cells() fails, the whole nvmem_register() will fail > and the cells will then be removed anyway. This is a prepartion s/prepartion/preparation/ Sascha
Am 2022-11-21 08:51, schrieb Sascha Hauer: > On Fri, Nov 18, 2022 at 07:51:17PM +0100, Michael Walle wrote: >> From: Miquel Raynal <miquel.raynal@bootlin.com> >> >> This layout applies no top of any non volatile storage device >> containing > > s/no/on/ > >> + table_len = hdr_len + data_len; >> + if (table_len > ONIE_TLV_MAX_LEN) { >> + dev_err(dev, "Invalid ONIE TLV data length\n"); >> + return -EINVAL; >> + } >> + >> + table = devm_kmalloc(dev, table_len, GFP_KERNEL); >> + if (!table) >> + return -ENOMEM; >> + >> + ret = nvmem_device_read(nvmem, 0, table_len, table); >> + if (ret != table_len) >> + goto free_data_buf; >> + >> + if (!onie_tlv_crc_is_valid(dev, table_len, table)) { >> + ret = -EINVAL; >> + goto free_data_buf; >> + } >> + >> + data = table + hdr_len; >> + ret = onie_tlv_add_cells(dev, nvmem, data_len, data); >> + if (ret) >> + goto free_data_buf; >> + >> +free_data_buf: >> + kfree(table); > > This is allocated with devm_kmalloc(), you shouldn't free it here. Miquel, should I drop your patches or should I fix this in place and repost them? Also, there is a mistake in nvmem_add_cells_from_layout() which doesn't handle the return code. Probably a leftover from before when I had .add_cells() return void. -michael