Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
2013/9/27 Richard Weinberger : > Am 27.09.2013 15:16, schrieb Richard Genoud: >> 2013/9/27 Richard Weinberger : >>> Am 27.09.2013 14:51, schrieb Richard Genoud: Some old UBI implementations (e.g. U-Boot) have not implemented the image sequence feature. So, when erase blocks are written, the image sequence in the ec header is lost (set to zero). UBI scan_all() takes this case into account (commits 32bc4820287a1a03982979515949e8ea56eac641 and 2eadaad67b2b6bd132eda105128d2d466298b8e3) But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. This patch fixes the issue. Signed-off-by: Richard Genoud --- >>> >>> W00t! Good catch! >>> >>> Acked-by: Richard Weinberger >> Thanks :) >> >> >> There's still an error when the image sequence is bad: >> [ 35.632812] UBI error: scan_pool: bad image seq: 0x0, expected: 0x6e452f03 >> [ 35.640625] UBI error: ubi_scan_fastmap: Attach by fastmap failed, >> doing a full scan! >> [ 35.648437] kmem_cache_destroy ubi_ainf_peb_slab: Slab cache still >> has objects <- the destroy_ai in line 1415 > > *grrr*, the problem here is that not all allocations which are done via > kmem_cache_alloc(ai->aeb_slab_cache, ...) got kfree()'d. > > Thanks, > //richard My first guess is in fastmap.c, function scan_pool(), in the loop "for (i = 0; i < pool_size; i++)" there's some "add_aeb()" and "new_aeb = kmem_cache_alloc()" And on error, I don't see a loop backward to free those ones. Richard. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
Am 27.09.2013 15:16, schrieb Richard Genoud: > 2013/9/27 Richard Weinberger : >> Am 27.09.2013 14:51, schrieb Richard Genoud: >>> Some old UBI implementations (e.g. U-Boot) have not implemented the image >>> sequence feature. >>> So, when erase blocks are written, the image sequence in the ec header >>> is lost (set to zero). >>> UBI scan_all() takes this case into account (commits >>> 32bc4820287a1a03982979515949e8ea56eac641 and >>> 2eadaad67b2b6bd132eda105128d2d466298b8e3) >>> >>> But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. >>> >>> This patch fixes the issue. >>> >>> Signed-off-by: Richard Genoud >>> --- >> >> W00t! Good catch! >> >> Acked-by: Richard Weinberger > Thanks :) > > > There's still an error when the image sequence is bad: > [ 35.632812] UBI error: scan_pool: bad image seq: 0x0, expected: 0x6e452f03 > [ 35.640625] UBI error: ubi_scan_fastmap: Attach by fastmap failed, > doing a full scan! > [ 35.648437] kmem_cache_destroy ubi_ainf_peb_slab: Slab cache still > has objects <- the destroy_ai in line 1415 *grrr*, the problem here is that not all allocations which are done via kmem_cache_alloc(ai->aeb_slab_cache, ...) got kfree()'d. Thanks, //richard -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
2013/9/27 Richard Weinberger : > Am 27.09.2013 14:51, schrieb Richard Genoud: >> Some old UBI implementations (e.g. U-Boot) have not implemented the image >> sequence feature. >> So, when erase blocks are written, the image sequence in the ec header >> is lost (set to zero). >> UBI scan_all() takes this case into account (commits >> 32bc4820287a1a03982979515949e8ea56eac641 and >> 2eadaad67b2b6bd132eda105128d2d466298b8e3) >> >> But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. >> >> This patch fixes the issue. >> >> Signed-off-by: Richard Genoud >> --- > > W00t! Good catch! > > Acked-by: Richard Weinberger Thanks :) There's still an error when the image sequence is bad: [ 35.632812] UBI error: scan_pool: bad image seq: 0x0, expected: 0x6e452f03 [ 35.640625] UBI error: ubi_scan_fastmap: Attach by fastmap failed, doing a full scan! [ 35.648437] kmem_cache_destroy ubi_ainf_peb_slab: Slab cache still has objects <- the destroy_ai in line 1415 [ 35.656250] CPU: 0 PID: 486 Comm: ubiattach Not tainted 3.11.1 #38 [ 35.656250] [] (unwind_backtrace+0x0/0xe0) from [] (show_stack+0x10/0x14) [ 35.671875] [] (show_stack+0x10/0x14) from [] (destroy_ai+0x230/0x244) [ 35.679687] [] (destroy_ai+0x230/0x244) from [] (ubi_attach+0x1f8/0x390) [ 35.687500] [] (ubi_attach+0x1f8/0x390) from [] (ubi_attach_mtd_dev+0x2b8/0x86c) [ 35.695312] [] (ubi_attach_mtd_dev+0x2b8/0x86c) from [] (ctrl_cdev_ioctl+0xd0/0x184) [ 35.703125] [] (ctrl_cdev_ioctl+0xd0/0x184) from [] (vfs_ioctl+0x28/0x3c) [ 35.710937] [] (vfs_ioctl+0x28/0x3c) from [] (do_vfs_ioctl+0x4e8/0x54c) [ 35.718750] [] (do_vfs_ioctl+0x4e8/0x54c) from [] (SyS_ioctl+0x34/0x58) [ 35.726562] [] (SyS_ioctl+0x34/0x58) from [] (ret_fast_syscall+0x0/0x2c) I'll look into that Richard -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
Am 27.09.2013 14:51, schrieb Richard Genoud: > Some old UBI implementations (e.g. U-Boot) have not implemented the image > sequence feature. > So, when erase blocks are written, the image sequence in the ec header > is lost (set to zero). > UBI scan_all() takes this case into account (commits > 32bc4820287a1a03982979515949e8ea56eac641 and > 2eadaad67b2b6bd132eda105128d2d466298b8e3) > > But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. > > This patch fixes the issue. > > Signed-off-by: Richard Genoud > --- W00t! Good catch! Acked-by: Richard Weinberger -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
Am 27.09.2013 14:51, schrieb Richard Genoud: Some old UBI implementations (e.g. U-Boot) have not implemented the image sequence feature. So, when erase blocks are written, the image sequence in the ec header is lost (set to zero). UBI scan_all() takes this case into account (commits 32bc4820287a1a03982979515949e8ea56eac641 and 2eadaad67b2b6bd132eda105128d2d466298b8e3) But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. This patch fixes the issue. Signed-off-by: Richard Genoud richard.gen...@gmail.com --- W00t! Good catch! Acked-by: Richard Weinberger rich...@nod.at -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
2013/9/27 Richard Weinberger rich...@nod.at: Am 27.09.2013 14:51, schrieb Richard Genoud: Some old UBI implementations (e.g. U-Boot) have not implemented the image sequence feature. So, when erase blocks are written, the image sequence in the ec header is lost (set to zero). UBI scan_all() takes this case into account (commits 32bc4820287a1a03982979515949e8ea56eac641 and 2eadaad67b2b6bd132eda105128d2d466298b8e3) But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. This patch fixes the issue. Signed-off-by: Richard Genoud richard.gen...@gmail.com --- W00t! Good catch! Acked-by: Richard Weinberger rich...@nod.at Thanks :) There's still an error when the image sequence is bad: [ 35.632812] UBI error: scan_pool: bad image seq: 0x0, expected: 0x6e452f03 [ 35.640625] UBI error: ubi_scan_fastmap: Attach by fastmap failed, doing a full scan! [ 35.648437] kmem_cache_destroy ubi_ainf_peb_slab: Slab cache still has objects - the destroy_ai in line 1415 [ 35.656250] CPU: 0 PID: 486 Comm: ubiattach Not tainted 3.11.1 #38 [ 35.656250] [c001283c] (unwind_backtrace+0x0/0xe0) from [c0010a24] (show_stack+0x10/0x14) [ 35.671875] [c0010a24] (show_stack+0x10/0x14) from [c019b9e4] (destroy_ai+0x230/0x244) [ 35.679687] [c019b9e4] (destroy_ai+0x230/0x244) from [c019d334] (ubi_attach+0x1f8/0x390) [ 35.687500] [c019d334] (ubi_attach+0x1f8/0x390) from [c0191a18] (ubi_attach_mtd_dev+0x2b8/0x86c) [ 35.695312] [c0191a18] (ubi_attach_mtd_dev+0x2b8/0x86c) from [c0192208] (ctrl_cdev_ioctl+0xd0/0x184) [ 35.703125] [c0192208] (ctrl_cdev_ioctl+0xd0/0x184) from [c007ef4c] (vfs_ioctl+0x28/0x3c) [ 35.710937] [c007ef4c] (vfs_ioctl+0x28/0x3c) from [c007f974] (do_vfs_ioctl+0x4e8/0x54c) [ 35.718750] [c007f974] (do_vfs_ioctl+0x4e8/0x54c) from [c007fa0c] (SyS_ioctl+0x34/0x58) [ 35.726562] [c007fa0c] (SyS_ioctl+0x34/0x58) from [c000e3c0] (ret_fast_syscall+0x0/0x2c) I'll look into that Richard -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
Am 27.09.2013 15:16, schrieb Richard Genoud: 2013/9/27 Richard Weinberger rich...@nod.at: Am 27.09.2013 14:51, schrieb Richard Genoud: Some old UBI implementations (e.g. U-Boot) have not implemented the image sequence feature. So, when erase blocks are written, the image sequence in the ec header is lost (set to zero). UBI scan_all() takes this case into account (commits 32bc4820287a1a03982979515949e8ea56eac641 and 2eadaad67b2b6bd132eda105128d2d466298b8e3) But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. This patch fixes the issue. Signed-off-by: Richard Genoud richard.gen...@gmail.com --- W00t! Good catch! Acked-by: Richard Weinberger rich...@nod.at Thanks :) There's still an error when the image sequence is bad: [ 35.632812] UBI error: scan_pool: bad image seq: 0x0, expected: 0x6e452f03 [ 35.640625] UBI error: ubi_scan_fastmap: Attach by fastmap failed, doing a full scan! [ 35.648437] kmem_cache_destroy ubi_ainf_peb_slab: Slab cache still has objects - the destroy_ai in line 1415 *grrr*, the problem here is that not all allocations which are done via kmem_cache_alloc(ai-aeb_slab_cache, ...) got kfree()'d. Thanks, //richard -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] UBI: fastmap: fix backward compatibility with image_seq
2013/9/27 Richard Weinberger rich...@nod.at: Am 27.09.2013 15:16, schrieb Richard Genoud: 2013/9/27 Richard Weinberger rich...@nod.at: Am 27.09.2013 14:51, schrieb Richard Genoud: Some old UBI implementations (e.g. U-Boot) have not implemented the image sequence feature. So, when erase blocks are written, the image sequence in the ec header is lost (set to zero). UBI scan_all() takes this case into account (commits 32bc4820287a1a03982979515949e8ea56eac641 and 2eadaad67b2b6bd132eda105128d2d466298b8e3) But fastmap scan functions (ubi_scan_fastmap() and scan_pool()) didn't. This patch fixes the issue. Signed-off-by: Richard Genoud richard.gen...@gmail.com --- W00t! Good catch! Acked-by: Richard Weinberger rich...@nod.at Thanks :) There's still an error when the image sequence is bad: [ 35.632812] UBI error: scan_pool: bad image seq: 0x0, expected: 0x6e452f03 [ 35.640625] UBI error: ubi_scan_fastmap: Attach by fastmap failed, doing a full scan! [ 35.648437] kmem_cache_destroy ubi_ainf_peb_slab: Slab cache still has objects - the destroy_ai in line 1415 *grrr*, the problem here is that not all allocations which are done via kmem_cache_alloc(ai-aeb_slab_cache, ...) got kfree()'d. Thanks, //richard My first guess is in fastmap.c, function scan_pool(), in the loop for (i = 0; i pool_size; i++) there's some add_aeb() and new_aeb = kmem_cache_alloc() And on error, I don't see a loop backward to free those ones. Richard. -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
